---
title: Huggins Vorticity Flux Tensor Study
url: https://www.emergentmind.com/topics/huggins-vorticity-flux-tensor
type: topic
---

# Huggins Vorticity Flux Tensor Study

Searching arXiv for recent papers on the Huggins vorticity-flux tensor and Josephson–Anderson relation.
arxiv_search(query="Huggins vorticity flux tensor Josephson Anderson drag", max_results=10, sort_by="relevance")
Attempting to retrieve relevant arXiv records.
The Huggins vorticity flux tensor, usually denoted \(\Sigma\), is the anti-symmetric tensor that appears in the Helmholtz equation for vorticity,
\[
\partial_t \omega + \nabla\cdot\Sigma = 0,
\]
and encodes the transport of vorticity by advection, vortex tilting and stretching, viscosity, and, in generalized formulations, non-conservative body-force. In recent analyses of viscous bluff-body flows, it serves as the local quantity that links vorticity dynamics to drag through the detailed Josephson–Anderson relation. That role is developed for flow over a sphere at \(Re=\{200,3700\}\) and over a prolate spheroid at \(Re=3000\) and \(20^\circ\) incidence, and is also extended to streamwise-periodic channel flows [2507.09096] [2407.01416].

## 1. Definition and equivalent representations

For the bluff-body flows over a sphere and a prolate spheroid, \(\Sigma\) is introduced in index form as
\[
\Sigma_{ij} = u_i\omega_j - \omega_i u_j - \nu(\partial_i\omega_j-\partial_j\omega_i),
\]
or, equivalently,
\[
\Sigma = u\,\omega^T - \omega\,u^T - \nu\bigl(\nabla\omega-(\nabla\omega)^T\bigr).
\]
The tensor is anti-symmetric, \(\Sigma_{ij}=-\Sigma_{ji}\), so it contains three independent components rather than nine [2507.09096].

The same study also introduces the dual axial-vector form
\[
\eta \equiv u\times\omega - \nu\nabla\times\omega,\qquad
\Sigma_{ij}=\epsilon_{ijk}\eta_k,\qquad
\eta_i=\tfrac12\epsilon_{ijk}\Sigma_{jk}.
\]
This representation is central in the drag formula because the Josephson–Anderson relation is written directly in terms of \(\eta\) as well as \(\Sigma\) [2507.09096].

In the streamwise-periodic channel formulation, the tensor is written more generally as
\[
\Sigma_{ij}
=
u_i\omega_j-u_j\omega_i
-\nu(\partial_i\omega_j-\partial_j\omega_i)
-\epsilon_{ijk}f_k,
\]
with \(\mathbf f\) any non-conservative body-force. In that formulation, \(\Sigma\) is sometimes denoted \(J_{ij}\) in other works [2407.01416].

## 2. Constituent fluxes and planar interpretation

A standard decomposition is
\[
\Sigma=\Sigma_a+\Sigma_\nu,
\]
with
\[
\Sigma_{a,ij}=u_i\omega_j-\omega_i u_j,
\qquad
\Sigma_{\nu,ij}=-\nu(\partial_i\omega_j-\partial_j\omega_i).
\]
The first term is the advective flux, which physically transports each component \(\omega_j\) of vorticity along the velocity \(u_i\) and re-orients it via tilting and stretching. The second term is the viscous flux, the anti-symmetric gradient of vorticity, and embodies down-gradient diffusion of vorticity by viscosity [2507.09096].

In the channel generalization, a third contribution appears when non-conservative body-force is present. The paper describes the convective term as a “Reynolds-stress” contribution and the body-force term as a Magnus-type contribution [2407.01416].

| Contribution | Expression | Interpretation |
|---|---|---|
| Advective flux | \(u_i\omega_j-\omega_i u_j\) | Transport plus tilting and stretching |
| Viscous flux | \(-\nu(\partial_i\omega_j-\partial_j\omega_i)\) | Down-gradient diffusion |
| Body-force flux | \(-\epsilon_{ijk}f_k\) | Magnus-type contribution |

For any fixed direction \(e_\xi\), the planar conservation law is
\[
\partial_t\omega_\xi+\nabla\cdot(\Sigma\cdot e_\xi)=0.
\]
Accordingly, \(\Sigma\cdot e_\xi\) is the instantaneous flux of the vorticity component \(\omega_\xi\) within the plane normal to \(e_\xi\). This formulation is used explicitly for azimuthal vorticity in the axisymmetric sphere and for streamwise vorticity in the prolate spheroid [2507.09096].

## 3. Detailed Josephson–Anderson relation

For an isolated bluff body held fixed in a uniform stream \(U\), the detailed Josephson–Anderson relation is written as
\[
-\,F\cdot U
=
\rho\int_\Omega u_\phi\cdot(u\times\omega-\nu\nabla\times\omega)\,dV
=
\rho\int_\Omega u_\phi\cdot\eta\,dV
=
-\tfrac12\int dJ\int \epsilon_{ijk}\Sigma_{ij}\,d\ell_k.
\]
Here \(u_\phi\) is the background potential-flow velocity, \(dJ=\rho|u_\phi|\,dA\) labels streamtubes of the potential flow, and \(d\ell_k\) is a line element along \(u_\phi\) [2507.09096].

Because \(\epsilon_{ijk}\Sigma_{ij}d\ell_k\) selects only those \(\Sigma_{ij}\) whose flux index \(i\) and vorticity index \(j\) are both orthogonal to the streamline direction \(k\), the relation equates the instantaneous drag-power \(-F\cdot U\) to the net rate at which streamwise mass currents cut through vorticity. In that interpretation, outward crossing of negative vorticity gives positive drag-power, whereas inward crossing gives anti-drag [2507.09096].

The periodic-channel extension preserves the same structure. For fixed mass flux \(J\), the channel paper obtains
\[
\mathcal T=(\Delta h)J=\gamma L_x J,
\]
and also rewrites \(\mathcal T\) as a surface integral of \(\Sigma_{ij}\) and, in a canonical channel with \(\mathbf u_\phi=U\hat{\mathbf x}\), as
\[
\mathcal T
=
-\rho U\int_\Omega \Sigma_{yz}\,dV
=
-\rho U\int_\Omega
\bigl[\omega_z v-\omega_y w-\nu\partial_y\omega_z\bigr]\,dV.
\]
In that setting, the total drag is the integrated wall-normal flux of spanwise vorticity [2407.01416].

## 4. Axisymmetric sphere at \(Re=200\)

For the laminar sphere, the flow is axisymmetric about \(x\), and only the azimuthal component \(\omega_\phi\) is nonzero. Using cylindrical coordinates \((x,r,\phi)\), the study defines local bases \((e_s,e_n,e_\phi)\) aligned with the streamfunction, its normal, and the azimuthal direction. The tensor is then decomposed as
\[
\Sigma_a =
\begin{bmatrix}
0 & 0 & u_x\omega_\phi\\
0 & 0 & u_r\omega_\phi\\
-u_x\omega_\phi & -u_r\omega_\phi & 0
\end{bmatrix},
\qquad
\Sigma_\nu =
\begin{bmatrix}
0 & 0 & -\nu\partial_x\omega_\phi\\
0 & 0 & -\nu\partial_r\omega_\phi\\
\nu\partial_x\omega_\phi & \nu\partial_r\omega_\phi & 0
\end{bmatrix}.
\]
Projecting onto \(e_\phi\) gives
\[
\Sigma_a\cdot e_\phi=(u_x\omega_\phi,\;u_r\omega_\phi,\;0)^T,
\qquad
\Sigma_\nu\cdot e_\phi=(-\nu\partial_x\omega_\phi,\;-\nu\partial_r\omega_\phi,\;0)^T.
\]
The corresponding scalar Josephson–Anderson integrands are
\[
\Pi_a(x)=|u_\phi|\,e_n\cdot\Sigma_a\cdot(-e_\phi),
\qquad
\Pi_\nu(x)=|u_\phi|\,e_n\cdot\Sigma_\nu\cdot(-e_\phi)
\]
[2507.09096].

The analysis identifies three distinct roles for the tensor components. First, the viscous flux near the wall drives \(\omega_\phi\) into the fluid and is responsible for the local drag. Second, the advective flux in the detached shear layer moves \(\omega_\phi\) outward across streamlines, producing drag, and later inward near the wake centerline, producing anti-drag. Third, viscous annihilation of azimuthal vorticity at the wake centerline contributes to pressure recovery. The paper states that the Josephson–Anderson relation is first demonstrated for the sphere at \(Re=200\) [2507.09096].

## 5. Impulsively started and turbulent sphere at \(Re=3700\)

For the sphere at \(Re=3700\), the tensor is evaluated both instantaneously during impulsive start and in the stationary turbulent regime. In the averaged turbulent wake, the decomposition is extended to
\[
\overline\Sigma
=
\overline\Sigma_{\overline a}
+
\overline\Sigma_{a'}
+
\overline\Sigma_\nu,
\]
where \(\overline\Sigma_{\overline a}\) is mean-flow advection of mean vorticity, \(\overline\Sigma_{a'}\) is the Reynolds-stress-like turbulent flux of vorticity, and \(\overline\Sigma_\nu\) is mean viscous diffusion. The corresponding projected Josephson–Anderson contributions are \(\overline\Pi_{\overline a}\), \(\overline\Pi_{a'}\), and \(\overline\Pi_\nu\), and the averaged drag relation is
\[
-\,\overline{F\cdot U}
=
-\rho\int dJ\int
e_n\cdot
\bigl(
\overline\Sigma_{\overline a}
+
\overline\Sigma_{a'}
+
\overline\Sigma_\nu
\bigr)
\cdot e_\phi\,d\ell
\]
[2507.09096].

The impulsive-start analysis assigns all of the early singular drag to wall diffusion via \(\Sigma_\nu\), rather than to a conventional pressure-plus-shear partition. As the boundary layer detaches and the wake turns turbulent, \(\Sigma_a\) grows and eventually dominates drag production, while \(\Sigma_\nu\) decays [2507.09096].

In the fully turbulent wake, mean advective flux in the shear layer dominates near the body, whereas the turbulent flux \(\overline\Pi_{a'}\) takes over near the centerline. The turbulent flux enhances the transport of mean azimuthal vorticity toward the wake centerline and is identified as the driver of enthalpy recovery downstream. The same mechanism is described as rapid pressure recovery, or anti-drag, farther downstream. Mean viscous flux is negligible except immediately at the wall and in the early wake [2507.09096].

## 6. Prolate spheroid at \(Re=3000\) and \(20^\circ\) incidence

For the prolate spheroid, the computed quantity is \(\Sigma\cdot e_x\), the instantaneous planar flux of streamwise vorticity \(\omega_x\) in vertical cuts. In this setting, \(\Sigma_a\cdot e_x\) shows advection of \(\omega_x\) along primary and secondary separated layers into the free stream, while \(\Sigma_\nu\cdot e_x\) shows diffusion of wall-generated \(\omega_x\). The integrated Josephson–Anderson contributions \(\Pi_a\) and \(\Pi_\nu\) are plotted on vertical slices at \(x/a=\{0.16,0.50,0.84\}\) [2507.09096].

The principal drag mechanism is transport of vorticity along the separated boundary layers. Advection of \(\omega_x\) in the primary separation, denoted \(B_1\) in the paper’s figure discussion, is the main driver of drag. The secondary layer, denoted \(B_2\), and the large-scale vortices redistribute vorticity and can locally reduce drag. The abstract states the same result more compactly: primary and secondary separation contribute oppositely to the drag force, while the large-scale vortices only re-distribute vorticity [2507.09096].

The viscous flux remains confined to the attached and immediately separated boundary layers, where it establishes the local wall-pressure gradients that drive the initial detachment. The study also proposes a mechanism for secondary separation based on the theory of vortex-induced separation [2507.09096].

## 7. Generalizations, scope, and boundary-condition correction

The periodic-channel study places the Huggins tensor in a broader setting. It states that the detailed Josephson–Anderson relation was first derived by Huggins for quantum superfluids, but also holds for internal flows of classical fluids and for external flows around solid bodies, corresponding there to relations of Burgers, Lighthill, Kambe, Howe and others. In all of these results, a background potential Euler flow with the same inflow and outflow as the physical flow is used, in the same spirit as Kelvin’s minimum energy theorem [2407.01416].

A technical issue arises in streamwise-periodic channels. The paper argues that Huggins’s original use of pure Neumann boundary conditions for the background potential creates an unphysical vortex sheet in a periodic channel, because only normal-velocity continuity is enforced and a tangential-velocity jump remains across the periodic join. The proposed correction is to impose Dirichlet conditions on \(\phi\) at inflow and outflow, together with Neumann conditions on the sidewalls. The resulting \(\mathbf u_\phi\) is \(x\)-periodic and \(C^\infty\), with no spurious vortex sheet; the orthogonality \(\int \mathbf u_\phi\cdot\mathbf u_\omega\,dV=0\) and the Josephson–Anderson derivation are preserved [2407.01416].

This extension reinforces the tensor’s role as a unified field quantity. In the bluff-body study, \(\Sigma\) and its projections \(\Pi\) identify where wall diffusion, shear-layer advection, turbulent mixing, and wake-centerline annihilation contribute positively or negatively to drag-power. In the periodic-channel study, the same tensor equates work by pressure drop to integrated spanwise-vorticity flux and reveals how vortex separation from a smooth bump creates drag at each time instant [2507.09096] [2407.01416].

Source: https://www.emergentmind.com/topics/huggins-vorticity-flux-tensor