Abstract
In this paper, an attempt is made to formulate a closure hypothesis for adverse pressure gradient turbulent boundary layers using the attached eddy hypothesis of Townsend and Perry & Chong, which was developed originally for zero pressure gradient layers and parallel duct flows. To the authors' knowledge, this work represents one of the few attempts to use coherent structure ideas in the formulation of a closure scheme. At present this closure scheme is primitive and many of the assumptions are of an arbituary nature but the analysis at least points out where the difficulties are and which areas need more work.