Recent efforts to develop and apply adaptive finite element techniques for
solving complex flow problems are reviewed. The emphasis is not on new meth
ods but on how to use existing methods to achieve accurate predictions. Var
ious sources of errors, and means of detecting them, are discussed. The aim
is to identify various sources of errors and their effects on adaptivity.
For simplicity and clarity, examples are taken from steady state two-dimens
ional flow problems. The paper discusses the use of adaptive techniques for
achieving grid-independent solutions. Wherever possible predictions are co
mpared with measurements. Copyright (C) 1999 John Wiley & Sons, Ltd.