# 8th Int'l Conference on Numerical Methods in Fluid Dynamics by Krause E. (ed.)

By Krause E. (ed.)

Similar computational mathematicsematics books

Numerical Methods for Structured Matrices and Applications: The Georg Heinig Memorial Volume

This cross-disciplinary quantity brings jointly theoretical mathematicians, engineers and numerical analysts and publishes surveys and learn articles on the topic of the themes the place Georg Heinig had made remarkable achievements. particularly, this comprises contributions from the fields of dependent matrices, quickly algorithms, operator concept, and purposes to approach concept and sign processing.

Robust Algebraic Multilevel Methods and Algorithms (Radon Series on Computational and Applied Mathematics)

This booklet offers with algorithms for the answer of linear structures of algebraic equations with large-scale sparse matrices, with a spotlight on difficulties which are acquired after discretization of partial differential equations utilizing finite aspect equipment. offers a scientific presentation of the new advances in powerful algebraic multilevel equipment.

Additional resources for 8th Int'l Conference on Numerical Methods in Fluid Dynamics

Example text

FIG. 13. Same as Fig. 51) is truncated after LN(x) with N = 40. Observe that the approximation is accurate for roughly 4% wavelengths of sin x. This page intentionally left blank SECTION 4 Review of Convergence Theory The fundamental problem of the numerical analysis of initial value problems is to find conditions under which UH(X, t) converges to u(x, /) as N"-* oo for some time interval 0 ^ ? ^ T and to estimate the error \\u — UN\\- The principal result is the Lax-Richtmyer equivalence theorem which states that stability is equivalent to convergence for consistent approximations to well-posed linear problems.

Laguerre expansions. (*) = 0, w(x) = e~x for 0 ^ x < o o with e~x/2n(x) bounded at x = 0 and oo. The nth eigenvalue is \n = n and the associated eigenfunction is (f>n(x) = Ln(x), the Laguerre polynomial of degree n. 30) that the Legendre expansion converges faster than algebraically as the number of terms JV-»oo. To illustrate the rate of convergence of Laguerre series, we consider the expansion of sin x: which converges for all x, 0 ^ x < oo. Since [see Erdelyi et al. (1953, p. 44*. 06polynomials per wavelength to achieve high accuracy.

Let us suppose that the Hilbert space 3€ possesses the inner product ( • , • ) • Using the inner product, we define (neglecting the complications due to boundary conditions) the adjoint L* of an operator L as that linear operator that satisfies (w, Lv)=(L*u, v) for all u, v in ffl. For the finite dimensional approximation LN, the matrix representation of L^ is the adjoint of the matrix representation of LN (see § 2). The operator LN is said to be a normal operator if LN commutes with Lfrso LNL\$,= L%LN.

Rated 4.98 of 5 – based on 9 votes

Categories: Computational Mathematicsematics