Numerical modelling of open channel junctions using the Riemann problem approach

TL;DR

This paper introduces a Riemann problem-based numerical model for open channel junctions that avoids empirical tuning and is validated against experimental, analytical, and two-dimensional data, showing improved accuracy especially with complex geometries.

Contribution

The paper presents a theoretically sound Riemann problem approach for open channel junctions that outperforms classic models, especially in complex geometrical scenarios.

Findings

Accurately models junction flows without empirical coefficients.

Performs well with complex geometries and unsteady conditions.

Validated against experimental, analytical, and 2D data.

Abstract

The solution of an extended Riemann problem is used to provide the internal boundary conditions at a junction when simulating one-dimensional flow through an open channel network. The proposed approach, compared to classic junction models, does not require the tuning of semi-empirical coefficients and it is theoretically well-founded. The Riemann problem approach is validated using experimental data, two-dimensional model results and analytical solutions. In particular, a set of experimental data is used to test each model under subcritical steady flow conditions, and different channel junctions are considered, with both continuous and discontinuous bottom elevation. Moreover, the numerical results are compared with analytical solutions in a star network to test unsteady conditions. Satisfactory results are obtained for all the simulations, and particularly for Y-shaped networks and for cases involving variations in channels' bottom and width. By contrast, classic models suffer when geometrical channel effects are involved.