Numerical and experimental investigations of a hydraulic pipe during a gate closure at a high Reynolds number

Paper in proceeding, 2015

The role of hydropower to provide regulated power is important to the Swedish power system. This becomes even more accentuated with the expansion of intermittent renewable electricity sources, such as wind power. The variation of hydropower operation ranges over a large spectrum of time scales, from seconds to years. For scales larger than a minute, the flow may be considered as quasi-steady from a hydrodynamic point of view. The present work addresses the shorter time scales. Such scales are manifested mainly as pressure transients, which is an issue of concern in design and operation of hydropower plants. The objective of the study is to address rapid pressure transients with a special focus on detailed 3D processes interacting with transients travelling in an essentially 1D geometry. The test case is a gate closing in a long rectangular pipe, where a high-Reynolds number flow is driven by a pressure difference between upper and lower water levels. Experimental time-resolved static pressure and PIV data are gathered for validation of the numerical results. In a first stage the computational domain is modelled in 3D with an incompressible volume of fluid method that includes the prediction of the free surfaces. The domain includes the upper and lower water tanks with free water surfaces, a pipe in-between and a closing and opening gate. The gate movement is modelled with a dynamic mesh that removes the cells as the gate closes. The block-structured mesh is generated in ICEM CFD, and parallel simulations are performed using the OpenFOAM open source software. The numerical results are compared with the experimental data, and it is shown that the experimentally observed pressure fluctuations after gate closure are not an effect of the free surfaces. In a second stage, the upper tank and the pipe are modelled using a compressible 1D code based on the method of characteristics (MOC). A comparison with the experimental data shows that the correct unsteady behavior of the system is captured by the 1D approach if the losses and the gate characteristics are correctly accounted for, at the same time as the compressibility is adapted to the air contents of the water and flexibility of the structure.