Curvature affects the flow of water in open channels in several ways. These effects include superelevationi secondary (indu ced) flows, longitudinal velocity redistribution, and bed scour in the case of a mobile boundary such as a natural stream or river. It is important for river engineers to understand and be able'to estimate these effects.

This study first presents and discusses the present state of knowledge and understanding of the problem of curved channel flow The flow mechanics are discussed first. withan emphasis op the prediction of specific flow phenomena. The state of the art is found to be adequate with regard to some effects but not others. A body of experimental data obtained under controlled conditions is an important component of the state of knowledge of a field. The existing curved channel experiments are tabulated and evaulated in the second part of this section. A discussion of the relatively recent application of numerical modelling is presented. Aspects of the problem reguiring further research and elaboration are also identified.

A theoretical analysis is then performed based on a perturbation of the governing flow equations. This analysis is confined to rectangular channels but may be extended to an arbitrary channel shape. The theoretical predictions of developed and developing longitudinal velocity distributions and maximums are found to agreе well with experimental measurements.

An experimental study comprising two complete runs ina singlelong (270°) bend was also performed. Use of the Laser Doppler Anaemometer tomeasure longitudinal and lateral velocity components allowed unprecedented detail and precision. Analysis of the velocity data provided the distributions of bed stress and bed stress angle which are important for scour prediction.

The purpose of this study was to apply recent research to practice and to attempt to fill a few of the gaps in the present state of knowledge of curved channel flows.