Smart algorithms and adaptive methods in computational fluid dynamics
Abstract
A review is presented of the use of smart algorithms which employ adaptive methods in processing large amounts of data in computational fluid dynamics (CFD). Smart algorithms use a rationally based set of criteria for automatic decision making in an attempt to produce optimal simulations of complex fluid dynamics problems. The information needed to make these decisions is not known beforehand and evolves in structure and form during the numerical solution of flow problems. Once the code makes a decision based on the available data, the structure of the data may change, and criteria may be reapplied in order to direct the analysis toward an acceptable end. Intelligent decisions are made by processing vast amounts of data that evolve unpredictably during the calculation. The basic components of adaptive methods and their application to complex problems of fluid dynamics are reviewed. The basic components of adaptive methods are: (1) data structures, that is what approaches are available for modifying data structures of an approximation so as to reduce errors; (2) error estimation, that is what techniques exist for estimating error evolution in a CFD calculation; and (3) solvers, what algorithms are available which can function in changing meshes. Numerical examples which demonstrate the viability of these approaches are presented.
 Publication:

Twelfth Canadian Congress of Applied Mechanics
 Pub Date:
 May 1989
 Bibcode:
 1989apme.proc...54T
 Keywords:

 Algorithms;
 Artificial Intelligence;
 Computational Fluid Dynamics;
 Computational Grids;
 Optimization;
 Problem Solving;
 Self Organizing Systems;
 Complex Systems;
 Correction;
 Data Processing;
 Data Structures;
 Decision Making;
 Error Correcting Codes;
 Fluid Mechanics and Heat Transfer