Diagnostic Algorithms for Contour Dynamics,
Abstract
The goal of large scale numerical simulations (numerical experiments) is to obtain a quantitative understanding of complicated nonlinear dynamical processes. A proper picture or graph can spark in-the prejudices of conservative intuitions. Diagnostic algorithms and their graphs are particularly useful in the contour dynamics model for studying two-dimensional fluid dynamics. This is because the 2D densities are replaced by contours bounding piecewise-constant density regions (i.e., 1D curves). Thus our diagnostic parameters are functions of one variable, the arc length along each curve, and their graphs are 2D. We discuss and illustrate some time-dependent properties of planar curves, including the spatial plot, low order moments, perimeter, curvature, and Fourier transforms. We also apply these techniques to contours obtained from finite-difference representations of continuum systems.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1984
- Accession Number
- ADP002947
Entities
People
- E. A. Overman Ii
- N. J. Zabusky
Organizations
- University of Pittsburgh