Clebsch representation

From HandWiki

In physics and mathematics, the Clebsch representation of an arbitrary three-dimensional vector field [math]\displaystyle{ \boldsymbol{v}(\boldsymbol{x}) }[/math] is:[1][2] [math]\displaystyle{ \boldsymbol{v} = \boldsymbol{\nabla} \varphi + \psi\, \boldsymbol{\nabla} \chi, }[/math]

where the scalar fields [math]\displaystyle{ \varphi(\boldsymbol{x}) }[/math][math]\displaystyle{ , \psi(\boldsymbol{x}) }[/math] and [math]\displaystyle{ \chi(\boldsymbol{x}) }[/math] are known as Clebsch potentials[3] or Monge potentials,[4] named after Alfred Clebsch (1833–1872) and Gaspard Monge (1746–1818), and [math]\displaystyle{ \boldsymbol{\nabla} }[/math] is the gradient operator.

Background

In fluid dynamics and plasma physics, the Clebsch representation provides a means to overcome the difficulties to describe an inviscid flow with non-zero vorticity – in the Eulerian reference frame – using Lagrangian mechanics and Hamiltonian mechanics.[5][6][7] At the critical point of such functionals the result is the Euler equations, a set of equations describing the fluid flow. Note that the mentioned difficulties do not arise when describing the flow through a variational principle in the Lagrangian reference frame. In case of surface gravity waves, the Clebsch representation leads to a rotational-flow form of Luke's variational principle.[8]

For the Clebsch representation to be possible, the vector field [math]\displaystyle{ \boldsymbol{v} }[/math] has (locally) to be bounded, continuous and sufficiently smooth. For global applicability [math]\displaystyle{ \boldsymbol{v} }[/math] has to decay fast enough towards infinity.[9] The Clebsch decomposition is not unique, and (two) additional constraints are necessary to uniquely define the Clebsch potentials.[1] Since [math]\displaystyle{ \psi\boldsymbol{\nabla}\chi }[/math] is in general not solenoidal, the Clebsch representation does not in general satisfy the Helmholtz decomposition.[10]

Vorticity

The vorticity [math]\displaystyle{ \boldsymbol{\omega}(\boldsymbol{x}) }[/math] is equal to[2]

[math]\displaystyle{ \boldsymbol{\omega} = \boldsymbol{\nabla}\times\boldsymbol{v} = \boldsymbol{\nabla}\times\left( \boldsymbol{\nabla} \varphi + \psi\, \boldsymbol{\nabla} \chi\right) = \boldsymbol{\nabla}\psi \times \boldsymbol{\nabla}\chi, }[/math]

with the last step due to the vector calculus identity [math]\displaystyle{ \boldsymbol{\nabla} \times (\psi \boldsymbol{A})=\psi(\boldsymbol{\nabla}\times\boldsymbol{A})+\boldsymbol{\nabla}\psi\times\boldsymbol{A}. }[/math] So the vorticity [math]\displaystyle{ \boldsymbol{\omega} }[/math] is perpendicular to both [math]\displaystyle{ \boldsymbol{\nabla}\psi }[/math] and [math]\displaystyle{ \boldsymbol{\nabla}\chi, }[/math] while further the vorticity does not depend on [math]\displaystyle{ \varphi. }[/math]

Notes

  1. Jump up to: 1.0 1.1 (Lamb 1993)
  2. Jump up to: 2.0 2.1 (Serrin 1959)
  3. (Benjamin 1984)
  4. (Aris 1962)
  5. (Clebsch 1859)
  6. (Bateman 1929)
  7. (Seliger Whitham)
  8. (Luke 1967)
  9. (Wesseling 2001)
  10. (Wu Ma)

References