CHAPTER 10

Vorticity & Circulation

Vorticity: The vorticity of a flow is defined as the curl of the velocity field:

vorticity : w
~ = ∇ × ~u
It is a microscopic measure of rotation (vector) at a given point in the fluid,
which can be envisioned by placing a paddle wheel into the flow. If it spins
about its axis at a rate Ω, then w = |w|
~ = 2Ω.

Circulation: The circulation around a closed contour C is defined as the

line integral of the velocity along that contour:
I Z
circulation : ΓC = ~u · d~l = w ~
~ · dS
C S

where S is an arbitrary surface bounded by C. The circulation is a macroscopic

measure of rotation (scalar) for a finite area of the fluid.

Irrotational fluid: An irrotational fluid is defined as being curl-free; hence,

w
~ = 0 and therefore ΓC = 0 for any C.

Vortex line: a line that points in the direction of the vortex vector. Hence
is vortex line is to w
~ what a streamlines is to ~u. Note that a vortex line asso-
ciated with a fluid line is always perpendicular to the streamline associated
with that fluid element.

Vortex tube: a bundle of vortex lines. The circularity of a curve C is

proportional to the number of vortex lines that thread the enclosed area.

In an inviscid fluid the vortex lines/tubes move with the fluid: a vortex line
anchored to some fluid element remains anchored to that fluid element.

53
Figure 3: Evolution of a vortex tube. Solid dots correspond to fluid elements.
Due to the shear in the velocity field, the vortex tube is stretched and tilted.
However, as long as the fluid is inviscid and barotropic, incompressible or
isobaric, Kelvin’s circularity theorem assures that the circularity is conserved
with time. In addition, since vorticity is divergence-free, the circularity along
different cross sections of the same vortex-tube is the same.

Vorticity equation: The Navier-Stokes momentum equations, in the ab-

sence of bulk viscosity, in Eulerian vector form, are given by
 
∂~u ∇P 2 1
+ (~u · ∇) ~u = − − ∇Φ + ν ∇ ~u + ∇(∇ · ~u)
∂t ρ 3
Using the vector identity (~u · ∇) ~u = 21 ∇u2 + (∇ × ~u) × ~u = ∇(u2 /2) − ~u × w
~
allows us to rewrite this as
 
∂~u ∇P 1 2 2 1
− ~u × w
~ =− − ∇Φ − ∇u + ν ∇ ~u + ∇(∇ · ~u)
∂t ρ 2 3
If we now take the curl on both sides of this equation, and we use that
~ = ∇2 (∇ × A),
curl(grad S) = 0 for any scalar field S, and that ∇ × (∇2 A) ~
we obtain the vorticity equation:

54
  
∂w
~ ∇P
= ∇ × (~u × w)
~ −∇ + ν∇2 w
~
∂t ρ

~ = ∇S × A
To write this in Lagrangian form, we first use that ∇ × (S A) ~+
~ [see Chapter 1] to write
S (∇ × A)

1 1 1 ρ∇(1) − 1∇ρ ∇P × ∇ρ
∇ ×( ∇P ) = ∇( ) ×∇P + (∇ ×∇P ) = ×∇P =
ρ ρ ρ ρ 2 ρ2
where we have used, once more, that curl(grad S) = 0. Next, using the vector
identities from Chapter 1, we write

∇ × (w
~ × ~u) = w(∇
~ · ~u) − (w
~ · ∇)~u − ~u(∇ · w)
~ + (~u · ∇)w
~
The third term vanishes because ∇ · w~ = ∇ · (∇ × ~u) = 0. Hence, using that
∂ w/∂t
~ − (~u · ∇)w
~ = dw/dt
~ we finally can write the vorticity equation in
Lagrangian form:

dw
~ ∇ρ × ∇P
= (w
~ · ∇)~u − w(∇
~ · ~u) + + ν∇2 w
~
dt ρ2

This equation describes how the vorticity of a fluid element evolves with
time. We now describe the various terms of the lhs of this equation in turn:

~ ·∇)~u: This term represents the stretching and tilting of vortex tubes
• (w
due to velocity gradients. To see this, we pick w
~ to be pointing in the
z-direction. Then

∂~u ∂ux ∂uy ∂uz

(w
~ · ∇)~u = wz = wz ~ex + wz ~ey + wz ~ez +
∂z ∂z ∂z ∂z
The first two terms on the rhs describe the tilting of the vortex tube,
while the third term describes the stretching.

• w(∇
~ · ~u): This term describes stretching of vortex tubes due to flow
compressibility. This term is zero for an incompressible fluid (∇·~u = 0).
Note that, again under the assumption that the vorticity is pointing in
the z-direction,

55
  
∂ux ∂uy ∂uz
w(∇
~ · ~u) = wz + + ~ez
∂x ∂y ∂z

• (∇ρ × ∇P )/ρ2 : This is the baroclinic term. It describes the production

of vorticity due to a misalignment between pressure and density gradi-
ents. This term is zero for a barotropic EoS: if P = P (ρ) the pressure
and density gradiens are parallel so that ∇P × ∇ρ = 0. Obviously,
this baroclinic also vanishes for an incompressible fluid (∇ρ = 0) or
for an isobaric fluid (∇P = 0). The baroclinic term is responsible, for
example, for creating vorticity in pyroclastic flows (see Figure 3).

• ν∇2 w:
~ This term describes the diffusion pf vorticity due to viscosity,
and is obvisouly zero for an inviscid fluid (ν = 0). Typically, viscosity
generates/creates vorticity at a bounding surface: due to the no-slip
boundary condition shear arises giving rise to vorticity, which is subse-
quently diffused into the fluid by the viscosity. In the interior of a fluid,
no new vorticity is generated; rather, viscosity diffuses and dissipates
vorticity.

• ∇ × F~ : There is a fifth term that can create vorticity, which however

does not appear in the vorticity equation above. The reason is that we
assumed that the only external force is gravity, which is a conservative
force and can therefore be written as the gradient of a (gravitational)
potential. More generally, though, there may be non-conservative, ex-
ternal body forces present, which would give rise to a ∇× F~ term in the
rhs of the vorticity equation. An example of a non-conservative force
creating vorticity is the Coriolis force, which is responsible for creating
hurricanes.

56
Figure 4: The baroclinic creation of vorticity in a pyroclastic flow. High
density fluid flows down a mountain and shoves itself under lower-density
material, thus creating non-zero baroclinicity.

Using the definition of circulation, one can shown that

Z  
dΓ ∂w
~ ~
= + ∇ × (w~ × ~u) · dS
dt S ∂t
Using the vorticity equation, this can be rewritten as

Z  
dΓ ∇ρ × ∇P 2 ~ ~
= + ν∇ w~ + ∇ × F · dS
dt S ρ2

where, for completeness, we have added in the contribution of an external

force F~ (which vanishes if F~ is conservative). Using Stokes’ Curl Theorem
we can also write this equation in a line-integral form as

dΓ ∇P ~
I I I
=− · dl + ν ∇ ~u · d~l +
2
F~ · d~l
dt ρ

which is the form that is more often used.

57
NOTE: By comparing the equations expressing dw/dt ~ and dΓ/dt it is clear
that the stretching a tilting terms present in the equation describing dw/dt,
~
are absent in the equation describing dΓ/dt. This implies that stretching
and tilting changes the vorticity, but keeps the circularity invariant. This is
basically the first theorem of Helmholtz described below.

Kelvin’s Circulation Theorem: The number of vortex lines that thread

any element of area that moves with the fluid remains unchanged in time
for an inviscid, barotropic (or incompressible) fluid, in the absence of non-
conservative forces.

The proof of Kelvin’s Circulation Theorem is immediately evident from

the above equation, which shows that dΓ/dt = 0 if the fluid is both inviscid
(ν = 0) and either barotropic (P = P (ρ) ⇒ ∇ρ × ∇P = 0) or incompressible
(∇ρ = 0), and there are no non-conservative forces (F~ = 0).

We end this chapter on vorticity and circulation with the three theorems of
Helmholtz, which hold in the absence of non-conservative forces (i.e., F~ = 0).

Helmholtz Theorem 1: The strength of a vortex tube, which is defined

as the circularity of the circumference of any cross section of the tube, is
constant along its length. This theorem holds for any fluid, and simply
derives from the fact that the vorticity field is divergence-free: ∇ · w~ =
∇ · (∇ × ~u) = 0. To see this, use Gauss’ divergence theorem to write that
Z Z
∇·w ~ dV = ~ · d2 S = 0
w
V S
Here V is the volume of a subsection of the vortex tube, and S is its bounding
surface. Since the vorticity is, by definition, perpendicular to S along the
sides of the tube, the only non-vanishing components to the surface integral
come from the areas at the top and bottom of the vortex tube; i.e.
Z Z Z
2~
w
~ ·d S = w
~ · (−n̂) dA + w
~ · n̂ dA = 0
S A1 A2

where A1 and A2 the areas of the cross sections of that bound the volume V
of the vortex tube. Using Stokes’ curl theorem, we have that

58
 Z I
w
~ · n̂ dA = ~u · d~l
A C
Hence we have that ΓC1 = ΓC2 where C1 and C2 are the curves bounding A1
and A2 , respectively.

Helmholtz Theorem 2: A vortex line cannot end in a fluid. Vortex lines

and tubes must appear as closed loops, extend to infinity, or start/end at
solid boundaries.

Helmholtz Theorem 3: A barotropic (or incompressible), inviscid fluid

that is initially irrotational will remain irrotational. Hence, such a fluid does
not and cannot create vorticity (except across curved shocks, see Chapter
11).
The proof of Helmholtz’ third theorem is straightforward. According to
Kelvin’s circulation theorem, a barotropic (or incompressible), inviscid fluid
has dΓ/dt = 0 everywhere. Using that
Z  
dΓ ∂w~ ~=0
= + ∇ × (w~ × ~u) · d2 S
dt S ∂t
Since this has to hold for any S, we have that ∂ w/∂t
~ = ∇ × (~u × w).
~ Hence,
if w
~ = 0 initially, the vorticity remains zero for ever.

Potential Flow: An implication of Helmholtz’ third theorem is that if an

inviscid fluid is incompressible (∇ · ~u = 0) and irrotational (∇ × ~u = 0), then
it will remain irrotational. Such a flow is called Potential Flow, and obeys
Laplace’s equation:

∇2 Φu = o
where Φu is called the velocity potential, defined according to ~u = ∇Φu .
Although there is no such thing as an inviscid liquid, viscosity typically only
manifests itself in thin boundary layers (where Kelvin’s circulation theorem
doesn’t apply). Outside of the boundary layer, flow is often accurately de-
scribed by potential flow.

59
Figure 5: A beluga whale demonstrating Kelvin’s circulation theorem and
Helmholtz’ second theorem by producing a closed vortex tube under water,
made out of air.

60