P-Laplacian Explained

In mathematics, the p-Laplacian, or the p-Laplace operator, is a quasilinear elliptic partial differential operator of 2nd order. It is a nonlinear generalization of the Laplace operator, where

p

is allowed to range over

1<p<infty

. It is written as

\Deltapu:=\nabla(|\nablau|p-2\nablau).

Where the

|\nablau|p-2

is defined as

|\nablau|p-2=\left[style\left(

\partialu
\partialx1

\right)2++\left(

\partialu
\partialxn

\right)2

p-2
2
\right]

In the special case when

p=2

, this operator reduces to the usual Laplacian.[1] In general solutions of equations involving the p-Laplacian do not have second order derivatives in classical sense, thus solutions to these equations have to be understood as weak solutions. For example, we say that a function u belonging to the Sobolev space

W1,p(\Omega)

is a weak solution of

\Deltapu=0in\Omega

if for every test function

\varphi\in

infty
C
0(\Omega)
we have

\int\Omega|\nablau|p-2\nablau\nabla\varphidx=0

where

denotes the standard scalar product.

Energy formulation

The weak solution of the p-Laplace equation with Dirichlet boundary conditions

\begin{cases} -\Deltapu=f&in\Omega\\ u=g&on\partial\Omega \end{cases}

in a domain

\Omega\subsetRN

is the minimizer of the energy functional

J(u)=

1
p

\int\Omega|\nablau|pdx-\int\Omegafudx

W1,p(\Omega)

satisfying the boundary conditions in the trace sense.[1] In the particular case

f=1,g=0

and

\Omega

is a ball of radius 1, the weak solution of the problem above can be explicitly computed and is given by

u(x)=C

p
p-1
\left(1-|x|

\right)

where

C

is a suitable constant depending on the dimension

N

and on

p

only. Observe that for

p>2

the solution is not twice differentiable in classical sense.

See also

Sources

Further reading

Notes and References

  1. Evans, pp 356.