Neumann boundary condition

From Wikipedia, the free encyclopedia

In mathematics, a Neumann boundary condition (named after Carl Neumann) imposed on an ordinary differential equation or a partial differential equation specifies the values the derivative of a solution is to take on the boundary of the domain.

In the case of an ordinary differential equation, for example such as

\frac{d^2y}{dx^2} + 3 y = 1

on the interval [0,1], the Neumann boundary condition takes the form

y'(0) = α1
y'(1) = α2

where α1 and α2 are given numbers.

For a partial differential equation on a domain

\Omega\subset R^n,

for example

\nabla^{2} y + y = 0

(\nabla^{2} denotes the Laplacian), the Neumann boundary condition takes the form

\frac{\partial y}{\partial \nu}(x) = f(x) \quad \forall x \in \partial\Omega.

Here, ν denotes the (typically exterior) normal to the boundary ∂Ω and f is a given scalar function. The normal derivative which shows up on the left-hand side is defined as

\frac{\partial y}{\partial \nu}(x)=\nabla y(x)\cdot \nu (x)

where ∇ is the gradient and the dot is the inner product.

[edit] See also

In other languages