A New Variational Method for Initial Value Problems, Using Piecewise Hermite Polynomial Spline Functions.

Abstract

A variational principle for a functional can be found which satisfies both the original system and its adjoint system. The variations of this functional give no boundary terms if the bilinear concomitant of the systems vanishes. For a second order time varying initial value problem, one can adjust the boundary conditions of the adjoint system in terms of the boundary conditions of the original system so that the bilinear concomitant is identically zero. An expression for the variation of the functional is derived which contains only the terms involving the variations of the adjoint variable and its derivative, but no variation of its second derivative. The variations of the adjoint variable and its derivative are found to be zeroes at the final conditions, just as the variations of the original variable and its derivative are zero at the starting (initial) conditions. This implies that we are able to solve the problem in one direction without worrying about the conditions at the other end as the initial value problem should be. the algorithm is much more simplified than in the past. An example is given to show the procedures of this new variational method. (Author)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 01, 1981
Accession Number
ADA104184

Entities

People

  • C. N. Shen
  • Julian J. Wu

Organizations

  • United States Army Armament Research, Development and Engineering Center

Tags

DTIC Thesaurus Topics

  • Boundaries
  • Classification
  • Computations
  • Differential Equations
  • Equations
  • Integrals
  • Materials
  • Mechanical Properties
  • Military Research
  • Nuclear Reactors
  • Numerical Analysis
  • Oscillators
  • Polynomials
  • Security
  • Stationary
  • Variational Methods
  • Variational Principles

Fields of Study

  • Mathematics

Readers

  • Atmospheric Science / Meteorology, specifically Wind Wave Turbulence.
  • Finite Element Method (FEM) for solving Partial Differential Equations (PDEs)