A fluid flow is classified locally by its vorticity.

A flow is rotational where

$$\boxed{\boldsymbol\omega=\nabla\times\mathbf v\ne\mathbf0},$$

and irrotational where

$$\boxed{\boldsymbol\omega=\mathbf0}.$$

The words refer to the infinitesimal rotation of fluid elements, not simply to whether streamlines look curved or circular.

Rotational flow

In solid-body rotation,

$$\mathbf v=\boldsymbol\Omega\times\mathbf r,$$

and

$$\boldsymbol\omega=2\boldsymbol\Omega.$$

The fluid is rotational because each small material element has nonzero local angular velocity.

Simple shear is also rotational. Even though its streamlines can be straight and parallel, its vorticity is nonzero. This is an important warning: curved streamlines are not the definition of rotational flow.

Irrotational flow

A uniform velocity field

$$\mathbf v=U\hat{\mathbf x}$$

has no spatial velocity gradients, so

$$\boldsymbol\omega=\mathbf0.$$

The flow is irrotational.

More generally, a velocity field may vary strongly in space and still have zero curl because the different velocity gradients cancel in the vorticity combination.

Curved motion can still be irrotational

Consider the planar field, defined away from the origin,

$$\boxed{ \mathbf v(x,y)= \left( -\frac{Ky}{x^2+y^2}, \frac{Kx}{x^2+y^2} \right) }.$$

Its streamlines are circles centered on the origin, and the tangential speed is

$$v_\theta=\frac{K}{r}.$$

Yet direct differentiation gives

$$\boxed{\boldsymbol\omega=\mathbf0}$$

at every point with

$$r>0.$$

Thus fluid particles can move around circular streamlines while the surrounding flow is locally irrotational.

The angular speed of an orbiting particle,

$$\frac{v_\theta}{r}=rac{K}{r^2},$$

is not the same quantity as the local rigid-body rotation rate of an infinitesimal fluid element.

Irrotational flow can still have circulation around a hole

For a circular contour of radius $R$ around the origin,

$$\Gamma =\oint_C\mathbf v\cdot d\mathbf r =\left(\frac{K}{R}\right)(2\pi R),$$

so

$$\boxed{\Gamma=2\pi K}.$$

This is nonzero even though

$$\boldsymbol\omega=0$$

for every regular point in the flow domain.

There is no contradiction with Stokes' theorem: the velocity field is undefined at the origin. Any disk spanning the circular contour contains that singular point, so the required smoothness over the full spanning surface fails.

This ideal free vortex is a useful reminder that local irrotationality and global circulation are not identical concepts.

Local versus global statements

The condition

$$\nabla\times\mathbf v=0$$

is local. In a smooth simply connected region, every closed curve can be spanned without crossing a hole or singularity, so Stokes' theorem makes the circulation around such closed curves zero.

On a domain containing holes or excluded singularities, a flow can be locally irrotational while retaining nonzero circulation around those holes.

The topology of the flow domain therefore matters when turning a local curl statement into a global circulation conclusion.

Why the distinction matters

Irrotational-flow models are especially useful outside thin viscous regions, wakes, or vortex cores, where vorticity may be weak even though the overall flow is not spatially uniform.

Rotational regions, by contrast, contain genuine local vorticity and require models that track how that vorticity is produced, transported, stretched, or diffused.

The classification provides the conceptual boundary between two major branches of fluid mechanics:

• potential-flow theory, which exploits irrotational regions;
• vorticity dynamics, which follows the evolution of rotational motion.