scispace - formally typeset
Journal ArticleDOI

Eigenwertschranken für Eigenwertaufgaben mit partiellen Differentialgleichungen

Friedrich Goerisch, +1 more
- 01 Jan 1985 - 
- Vol. 65, Iss: 3, pp 129-135
Reads0
Chats0
TLDR
In this paper, a new procedure for calculating lower bounds for the eigenvalues is proposed, which is closely related to one developed by N. J. Lehmann, but it is essentially simpler to apply the new procedure to eigenvalue problems Mφ = ΛNφ, where M and N are partial differential operators.
Abstract
Betrachtet werden Eigenwertaufgaben Mφ = ΛNφ, wobei M und N symmetrische lineare Operatoren in einem Prahilbertraum sind. Zur Berechnung von unteren Schranken fur die Eigenwerte wird ein neues Verfahren vorgeschlagen. Es wird an Hand der Aufgabe erlautert, die bei der Bestimmung von Beulwerten eingespannter rechteckiger Platten auftritt; fur den kleinsten positiven Eigenwert werden genaue untere und obere Schranken angegeben. Das vorgeschlagene Verfahren ist eng mit einer von N. J. Lehmann entwickelten Methode verwandt, es kann aber sehr viel leichter auf Eigenwertaufgaben Mφ = ΛNφ angewandt werden, bei denen M und N partielle Differentialoperatoren sind. In the present paper eigenvalue problems of the type Mφ = ΛNφ are considered, where M and N are linear symmetric operators in a pre-Hilbert space. A new procedure for calculating lower bounds for the eigenvalues is proposed. The procedure is explained by means of the problem occurring in the determination of the buckling values for built-in rectangular plates. For the smallest positive eigenvalue accurate lower and upper bounds are given. The proposed procedure is closely related to one developed by N. J. Lehmann, but it is essentially simpler to apply the new procedure to eigenvalue problems Mφ = ΛNφ, where M and N are partial differential operators.

read more

Citations
More filters
Journal ArticleDOI

Lower Bounds for Eigenvalues of Elliptic Operators: By Nonconforming Finite Element Methods

TL;DR: A new systematic method that can produce lower bounds for eigenvalues is introduced and the saturation condition for most nonconforming elements is proved, which provides a guidance to modify known non Conforming elements in literature and to propose new nonconform elements.
Journal ArticleDOI

Eigenvalue inclusions for second-order ordinary differential operators by a numerical homotopy method

TL;DR: In this article, the inclusion intervals for the firstN eigenvalues of a second-order ordinary differential operator with boundary conditions of Sturm-Liouville or of periodic type are derived by a combination of "elementary" estimates, an appropriate numerical procedure and a homotopy algorithm.
Journal ArticleDOI

Bounds for eigenvalues of second-order elliptic differential operators

TL;DR: In this article, the firstN eigenvalues of a linear second-order elliptic differential operator on a bounded domain were derived by a combination of (a generalized version of) Kato's estimates and a homotopy algorithm.
Book ChapterDOI

Inclusion of eigenvalues of general eigenvalue problems for matrices

TL;DR: In this article, a procedure for calculating bounds to eigenvalues of general matrix eigenvalue problems is proposed, based on Temple quotients and their generalization by Lehmann, in connection with interval arithmetic.
Book ChapterDOI

Ein Stufenverfahren zur Berechnung von Eigenwertschranken

TL;DR: In this paper, a new way of applying inclusion theorems is proposed; it provides good bounds even in those cases where the usual procedure breaks down in the case where the procedure usually yields good bounds, but it sometimes fails.
References
More filters
Journal ArticleDOI

The Theory of Rayleigh's Principle as Applied to Continuous Systems

TL;DR: In this article, the Lagrangian method was extended to the problem of computing the frequency of the gravest mode of a vibrating system, and a series of successive approximations to the accurate solutions of the problems proposed were examined and a method was devised for obtaining an upper bound to the error involved in the approximate values of the frequency.