On the Evaluation of Bessel Functions

Abstract

In the present paper we describe an algorithm for the evaluation of Bessel functions J,(x), Yv(x) and H(j) (x) (j = 1, 2) of arbitrary positive orders and arguments at a constant CPU time. The algorithm employs Taylor series, the Debye asymptotic expansions and numerical evaluation of the Sommerfeld integral, and is based on the following two observations. (1) The Debye asymptotic expansions, contrary to what appears to be a popular belief, are not expansions in inverse powers of (large) parameter v but turn out to be uniform expansions in inverse powers of (large) parameter g1 = (x - V)/x1/3 for x > v and (large) parameter 92 = (v - x)/v1/3 for x < v. 2) For x and v such that both Taylor and Debye expansions do not provide a specified accuracy Bessel functions can be computed at a constant CPU time via (numerical) evaluation of the Sommerfeld integral along contours of steepest descents. In addition, in Appendix B we obtain certain new estimates concerning decay of the functions Jv(x) and -11Yv(x) of fixed x and large v, and in Appendix C we show that functions Jv(x) of integer v provide the solution for a certain system of coupled harmonic oscillators.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 08, 1992
Accession Number
ADA253129

Entities

People

  • Gregory Matviyenko

Organizations

  • Yale University

Tags

Communities of Interest

  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Analytic Functions
  • Asymptotic Series
  • Bessel Functions
  • Cancellation
  • Computations
  • Computer Science
  • Contour Integrals
  • Differential Equations
  • Equations
  • Errors
  • Inequalities
  • Integrals
  • Numerical Integration
  • Polynomials
  • Precision

Fields of Study

  • Mathematics

Readers

  • Finite Element Method (FEM) for solving Partial Differential Equations (PDEs)
  • Linear Algebra
  • Statistical inference.