|
Abstract
|
This study presents a new way to predict bedforms in open-channel flows by analyzing 811 laboratory experiments. Previous methods by Van Rijn (J Hydraulic Eng 110:1431–1456, 1984), Engelund (J Fluid Mechan 42:225–244, 1970), Simons and Richardson (1966), and Kennedy (J Fluid Mechanics 16:521–544, 1963) had limited success, with prediction accuracies of 77%, 76%, 72%, and 70%, respectively. These methods struggled with transitional flow regimes and various channel shapes. To fix these problems, a new dimensionless parameter (Φ) is extracted, which combines the Froude number, Shields parameter, relative roughness, and channel aspect ratio. The analysis showed strong links, including 0.82 between the Froude number and Shields parameter (Fr > 0.8, τ* > 0.5). The Φ-method got better prediction rates: 92% for lower-regime plane beds, 89% for ripples, 87% for dunes, 84% for transitional forms, and up to 88% for upper-regime bedforms. It performed robustly across flow conditions, with 91% accuracy in low-flow, fine-sediment scenarios (d₅₀ < 0.2 mm) and 80% in high-flow, coarse-sediment conditions (d₅₀ > 0.5 mm). The method excelled in both subcritical (88%) and supercritical (84%) regimes, offering a reliable framework for predicting bedform transitions in alluvial channels, especially where traditional methods falter.
|