Goodman and Kruskal's lambda: Difference between revisions
en>Melcombe →References: jstor |
en>Kku m l |
||
| Line 1: | Line 1: | ||
In mathematics, '''Neumann–Neumann methods''' are domain decomposition [[preconditioner]]s named so because they solve a [[Neumann problem]] on each subdomain on both sides of the interface between the subdomains.<ref name="Klawonn-2001-FNN">A. Klawonn and O. B. Widlund, ''FETI and Neumann–Neumann iterative substructuring methods: connections and new results'', Comm. Pure Appl. Math., 54 (2001), pp. 57–90.</ref> Just like all domain decomposition methods, so that the number of iterations does not grow with the number of subdomains, Neumann–Neumann methods require the solution of a coarse problem to provide global communication. The [[balancing domain decomposition]] is a Neumann–Neumann method with a special kind of coarse problem. | |||
More specifically, consider a domain Ω, on which we wish to solve the Poisson equation | |||
:<math>-\Delta u = f, \qquad u|_{\partial\Omega} = 0</math> | |||
for some function ''f''. Split the domain into two non-overlapping subdomains Ω<sub>1</sub> and Ω<sub>2</sub> with common boundary Γ and let ''u''<sub>1</sub> and ''u''<sub>2</sub> be the values of ''u'' in each subdomain. At the interface between the two subdomains, the two solutions must satisfy the matching conditions | |||
:<math>u_1 = u_2, \qquad \partial_nu_1 = \partial_nu_2</math> | |||
where ''n'' is the unit normal vector to Γ. | |||
An iterative method for approximating each u<sub>i</sub> satisfying the matching conditions is to first solve the decoupled problems (i=1,2) | |||
:<math>-\Delta u_i^{(k)} = f_i, \qquad u_i^{(k)}|_{\partial\Omega} = 0, \quad u^{(k)}_i|_\Gamma = \lambda^{(k)}</math> | |||
for some function λ<sup>(k)</sup> on Γ. We then solve the two Neumann problems | |||
:<math>-\Delta\psi_i^{(k)} = 0, \qquad \psi_i^{(k)}|_{\partial\Omega} = 0, \quad \partial_n\psi_i^{(k)} = \partial_nu_1^{(k)} - \partial_nu_2^{(k)}.</math> | |||
We then obtain the next iterate by setting | |||
:<math>\lambda^{(k+1)} = \lambda^{(k)} - \omega(\theta_1\psi_1^{(k)}|_\Gamma - \theta_2\psi_2^{(k)}|_\Gamma)</math> | |||
for some parameters ω, θ<sub>1</sub> and θ<sub>2</sub>. | |||
This procedure can be viewed as a [[Richardson extrapolation]] for the iterative solution of the equations arising from the [[Schur complement method]].<ref name="Quarteroni">A. Quarteroni and A. Valli, ''Domain Decomposition Methods for Partial Differential Equations'', Oxford Science Publications 1999.</ref> | |||
This continuous iteration can be discretized by the finite element method and then solved—in parallel—on a computer. The extension to more subdomains is straightforward, but using this method as stated as a preconditioner for the Schur complement system is not scalable with the number of subdomains; hence the need for a global coarse solve. | |||
==See also== | |||
* [[Neumann–Dirichlet method]] | |||
==References== | |||
<references/> | |||
{{Numerical PDE}} | |||
{{DEFAULTSORT:Neumann-Neumann Methods}} | |||
[[Category:Domain decomposition methods]] | |||
Revision as of 12:03, 10 January 2014
In mathematics, Neumann–Neumann methods are domain decomposition preconditioners named so because they solve a Neumann problem on each subdomain on both sides of the interface between the subdomains.[1] Just like all domain decomposition methods, so that the number of iterations does not grow with the number of subdomains, Neumann–Neumann methods require the solution of a coarse problem to provide global communication. The balancing domain decomposition is a Neumann–Neumann method with a special kind of coarse problem.
More specifically, consider a domain Ω, on which we wish to solve the Poisson equation
for some function f. Split the domain into two non-overlapping subdomains Ω1 and Ω2 with common boundary Γ and let u1 and u2 be the values of u in each subdomain. At the interface between the two subdomains, the two solutions must satisfy the matching conditions
where n is the unit normal vector to Γ.
An iterative method for approximating each ui satisfying the matching conditions is to first solve the decoupled problems (i=1,2)
for some function λ(k) on Γ. We then solve the two Neumann problems
We then obtain the next iterate by setting
for some parameters ω, θ1 and θ2.
This procedure can be viewed as a Richardson extrapolation for the iterative solution of the equations arising from the Schur complement method.[2]
This continuous iteration can be discretized by the finite element method and then solved—in parallel—on a computer. The extension to more subdomains is straightforward, but using this method as stated as a preconditioner for the Schur complement system is not scalable with the number of subdomains; hence the need for a global coarse solve.