Discharge prediction in straight mobile bed compound channels

Abstract

Most of the recent advances in techniques to predict the discharge capacity of compound channels have been based on experimental results from fixed bed laboratory models. This paper represents a significant departure from this trend by introducing a mobile bed in the main channel. As a result, sediment will be transported and reshaped by the flow in the main channel and this will lead to an increase in roughness as bedforms are created. The introduction of a mobile bed provides another degree of freedom and increases the complexity of the experiments. The discharge prediction problem still remains and methods for predicting the channel discharge capacity are examined. Experimental results from the United Kingdom Flood Channel Facility (UK FCF) and from the University of Ulster show that an improved estimate of both the zonal and overall discharge can be estimated using the Weighted Divided Channel method. Percentage errors for each discharge assessment were averaged for each data set; the most accurate method was the WDCM with an average error of 2·4 and the least accurate was the Divided Channel Method with vertical divisions with an over prediction in discharge of 16·0%. Results from tests on the rare situation of a compound channel with smooth floodplains and a rough main channel indicated that the usual explanation given for the effects of the momentum interaction between these two regions may need further investigation.