Solenoidal field

A vector field $\mathbf{F}$ is solenoidal (or divergence-free, source-free) if its divergence vanishes everywhere in its domain:

$\nabla \cdot \mathbf{F}=0.$

Equivalent characterisations on a simply connected domain:

1. $\nabla \cdot \mathbf{F}=0$ everywhere.
2. The flux of $\mathbf{F}$ through every closed surface is zero: $\iint \mathbf{F}\cdot d\mathbf{s}=0$.
3. $\mathbf{F}$ has a vector potential $\mathbf{A}$ such that $\mathbf{F}=\nabla \times \mathbf{A}$.

Solenoidal is the dual concept to conservative (curl-free, $\nabla \times \mathbf{F}=0$, equivalent to having a scalar potential). Conservative fields have scalar potentials and zero closed-loop circulation; solenoidal fields have vector potentials and zero closed-surface flux.

Why “solenoidal”

A solenoid is a coiled wire with current flowing through it; the field it produces is the prototypical solenoidal field. The current’s magnetic field $\mathbf{B}$ wraps around the coil but never flows out as a net source — flux into one end of the coil equals flux out the other. That’s the field-line behaviour the term captures: closed loops, no sources or sinks.

The “incompressible field” name comes from fluid mechanics: an incompressible fluid has $\nabla \cdot \mathbf{v}=0$ for its velocity field — fluid flows in must equal fluid flows out of every infinitesimal volume.

Why the three characterisations are equivalent

(1) ⇒ (2) by the Divergence theorem: $\iint \mathbf{F}\cdot d\mathbf{s}={\iiint }_V\nabla \cdot \mathbf{F}dv$. Zero divergence everywhere means zero volume integral, hence zero closed-surface flux.

(2) ⇒ (1): if some point had $\nabla \cdot \mathbf{F}\ne 0$, a small enclosing surface around it would have nonzero flux by continuity — contradicting (2).

(3) ⇒ (1): by the vector identity $\nabla \cdot (\nabla \times \mathbf{A})=0$. The divergence of any curl is identically zero; this is one of the standard divergence identities.

(1) ⇒ (3) is the deeper implication. On a simply connected (or, more precisely, contractible) domain, the Helmholtz theorem guarantees that any sufficiently smooth divergence-free field can be written as a curl. The construction of $\mathbf{A}$ involves integrating $\mathbf{F}$ over paths in a specific way, with a gauge freedom ($\mathbf{A}$ can be replaced by $\mathbf{A}+\nabla \chi$ for any scalar $\chi$ without changing $\nabla \times \mathbf{A}$).

On non-simply-connected domains, the implication can fail in the same way that conservative fields can fail to have global potentials — see Conservative vector field for the analogous obstruction.

Examples

Magnetic field $\mathbf{B}$. Always solenoidal: $\nabla \cdot \mathbf{B}=0$ (Gauss’s law for magnetism — no magnetic monopoles). Vector potential $\mathbf{A}$ is the Vector magnetic potential; $\mathbf{B}=\nabla \times \mathbf{A}$.

Velocity field of an incompressible fluid. $\nabla \cdot \mathbf{v}=0$. Vector potential is the Stream function in 2D (or its 3D vector analog).

Curl of any vector field. $\nabla \times \mathbf{G}$ is solenoidal for any $\mathbf{G}$, by $\nabla \cdot (\nabla \times \mathbf{G})=0$. So all curls are solenoidal — but not all solenoidal fields are obviously a curl until you construct $\mathbf{A}$.

Inverse-square radial field outside the source. $\mathbf{F}=\hat{\mathbf{r}}/r^2$ has $\nabla \cdot \mathbf{F}=0$ on ${\mathbb{R}}^3\setminus \{0\}$ — divergence-free everywhere except the singular origin. This is the magnetic-potential analog of the famous 2D rotational $\mathbf{F}=⟨-y,x⟩/(x^2+y^2)$ from the conservative-fields note: a globally-defined “almost solenoidal” field whose flux through any sphere enclosing the origin is $4\pi$, not zero. The non-trivial topology of ${\mathbb{R}}^3\setminus \{0\}$ is the obstruction.

Helmholtz decomposition

Every sufficiently smooth vector field $\mathbf{F}$ on a sufficiently nice domain can be uniquely decomposed as

$\mathbf{F}={\mathbf{F}}_{\text{conservative}}+{\mathbf{F}}_{\text{solenoidal}}=-\nabla \varphi +\nabla \times \mathbf{A},$

where ${\mathbf{F}}_{\text{conservative}}$ is curl-free and ${\mathbf{F}}_{\text{solenoidal}}$ is divergence-free. This Helmholtz decomposition is foundational in fluid dynamics and electromagnetism — it lets you separately analyse the “irrotational” and “incompressible” parts of any field.

In EM, the decomposition surfaces in the Helmholtz equation for the magnetic vector potential and in the analytical separation of longitudinal and transverse modes.

Solenoidal vs conservative — duality table

Property	Conservative	Solenoidal
Local condition	$\nabla \times \mathbf{F}=0$	$\nabla \cdot \mathbf{F}=0$
Closed-loop integral	${\oint }_C\mathbf{F}\cdot d\mathbf{r}=0$	$\iint \mathbf{F}\cdot d\mathbf{s}=0$
Potential	scalar $\varphi$, $\mathbf{F}=\nabla \varphi$	vector $\mathbf{A}$, $\mathbf{F}=\nabla \times \mathbf{A}$
Identity proving existence	$\nabla \times \nabla \varphi =0$	$\nabla \cdot (\nabla \times \mathbf{A})=0$
Flow of “stuff” picture	No vortices	No sources/sinks
Example	Gravitational field	Magnetic field
Topological obstruction	Loops around removed points	Closed surfaces around removed points

This duality is the geometric content of the de Rham complex in algebraic topology, but for engineering purposes it suffices to know: curl-free fields have scalar potentials, divergence-free fields have vector potentials, and both characterisations break down in the same kind of way when the domain has non-trivial topology.

In context

Solenoidal fields are central to magnetostatics ([[Magnetic flux density|$\mathbf{B}$ field]]), incompressible fluid mechanics, and the construction of vector potentials. The 2D special case is the Stream function. The dual concept is Conservative vector field. The decomposition into conservative + solenoidal parts is Helmholtz’s theorem — for the relevant identities see Curl and Divergence.