Press Room - Designing and maintaining weir structures

	Designing and maintaining weir structures
	Operational safety of hydraulic structures depends on the stability of the foundations, which need to be properly reinforced to avoid collapse. For example, the design of weir structures has in the past required time consuming and expensive physical testing across the complete range of flow conditions to which the structure may be exposed. Numerical simulations are an obvious way to optimise the design process, and consequently there is a need for validation. Weirs are used to regulate waterflow in streams and rivers (Figure 1). Liftable gates are placed between piers and the foundations of the structure are reinforced to avoid the effects of erosion. The flow through the weir is characterised as subcritical or supercritical. When conditions allow, a transition from supercritical to subcritical flow occurs and involves the dissipation of energy across an abrupt change in water depth, a ‘hydraulic jump’, and this may substantially affect the stability of the foundation. The so-called 'roller' at the hydraulic jump can be of two types. The first is a free-surface roller, which confines flow recirculation near the water surface at the hydraulic jump and is more erosive. The second is a surface roller covered from the downstream water which appears immediately downstream of the weir gate. This results in a higher water depth, and in this case the re-circulation occurs over the entire depth of water with reduced erosion effects. The question arises as to how far it is possible to calculate such a complex flow situation using numerical simulations and still make time and cost savings.
	To answer this question, physical model tests were carried out in a flow channel (2200 cm long, 100 cm wide and 100 cm deep) at the University of Hannover. A liftable gate with a thickness of 1 cm was located in the middle of the flow channel. By opening the gate at a height of 3.3 cm in a constant stream of 70.5 l/s, it was possible to investigate both types of hydraulic jump using two different loadings.
	Detailed measurements were taken of the water heights and the velocity distributions at 88 points on two measurement sections, 63 cm upstream and 165 cm downstream of the gate, using a 3D Acoustic Doppler Velocimeter flow probe (Figure 1). To complement the physical model tests, numerical calculations were run on the university's Silicon Graphics Origin Workstation using STAR-CD. The numerical model was confined to two dimensions, consisting of 5000 cells (Figure 2) and using the Volume-of Fluid-method. Furthermore, the performance of a variety of high Reynolds number turbulence models was investigated, namely the standard k/E model, the k/E-Chen model and the k/E Quadratic non-linear model. In conclusion, comparing the numerical results with the physical tests, the prediction of the water surface heights differs by no more than 6 % for both loading cases. Similarly, the flow velocity predictions do not differ by more than 6.5 % from the measurements. The effect of changing the turbulence model on the results and on the calculation times were marginal. Consequently, STAR-CD provides a new medium for designing and maintaining weir structures and can supplement the hitherto time-consuming and costly physical model tests carried out almost exclusively in this field (Figures 2 and 3). For further information please contact: tobilin@web.de (Tobias Linke), scheffermann@web.de (Jens Scheffermann) or spekker@web.de (Heiko Spekker).