R. Logar1, R. Klasinc2, F.
Steinman3
1Department for Hydraulic Maschinery, TU Graz
2 Department of Hydraulic Structures and Water Resources Management, TU Graz
3 Civil Engineering Department, Fluid Mechanics Laboratory, University Ljubljana
E-mail:klasinc@kwb.tu-graz.ac.at Tel.:+43316 873 8364 Fax.:+43316 873 8356
Abstract：Outlet and maintenance valves were
studied for a reservoir mainly used for irrigation. As relatively high flow
velocities were expected and as it had to be possible to close the maintenance
valve under full load, the valves were subjected to a hydraulic scale-model
test. Two designs were tested, a solid disk butterfly valve and a through-flow
version. The water supply at the laboratory was not sufficient to fulfill the
required pressure at the entrance to the model according to Froude’s law of
similitude. Therefore Euler’s law of similitude was used to determine the flow
and pressure relationship between prototype and model. By means of a simple
example the authors demonstrate that the flows determined by Froude’s
respectively Euler’s law agree very well in a wide range of head variation.
Keywords: reservoir, maintenance valve, hollow-cone valve, butterfly valve, disk version, through-flow version, hydraulic model, euler and froude similarity law
An irrigation reservoir has a bottom outlet consisting of two components: a feed gallery and a dividing system with three legs. Each leg ends in a hollow-cone valve, which cares for the greater part of energy dissipation. Upstream of each valve is an maintenance valve which in an emergency must allow closing even under full load. Two butterfly valve designs were to be considered, a solid disk valve and a through-flow valve (Fig. 1). Both versions were subjected to hydraulic scale model tests.
Fig. 1 Plan view of the model
The bottom outlet unit including dividing system and valves was reproduced by a model (scale 1:10). For test purposes, the first leg only was provided with a valve system, while the valves in the two other legs served for flow control only. The model was equipped with a multitude of measuring taps (Fig. 2) for recording static, and partly also dynamic pressures under different operating conditions.
Fig. 2 Location of pressure taps on the model
A set of measuring taps downstream of the maintenance valve is shown in Fig. 3. Measuring cross section K (4 pressure transducers with front membranes) and measuring taps L (10 measuring taps along the pipe) supplied data on presssure conditions downstream of the maintenance valve.
Fig. 3 Set of measuring taps (K,L and J)
First a solid disk butterfly valve was tested (Fig. 4). The following tests using the through-flow valve (Fig. 5) resulted in no major differences for the pressure readings both for static and dynamic pressures, except for the data supplied by the measuring taps of cross section L, which showed pressure fluctuations downstream of the valve to die out faster for the through-flow valve than for the solid disk valve (Fig. 6 and 7).
Fig. 4 Solid disk valve
Fig. 5 Through-flow valve
Fig. 6 Pressure fluctuations (taps L) Fig. 7 Pressure fluctuations downstream of
downstream of the solid disk valve the through-flow valve
Operational condition (Fig.6 and fig.7): Butterfly valve 45° ,hollow cone valve 100%
Froude’s laws of similitude are sufficient for our model analyses even if the pressure heads do not correspond to the necessary values. With an available head in the reservoir of 100 m and a head loss of approx. 10 m in the conduit a head of 90 m is available at the entrance to the bifurcation. This head - converted according to the Froude law - corresponds to a head of 9m , or a pressure of 0.9 bar, on the model (model scale, 1:10). It is generally known that Euler’s model laws are not tied to this strict rule of Froude’s law. Euler’s laws apply independently of the entrance pressure. In our investigations a pressure of only 0.4 bar was available.
Using a simple example, it was possible to demonstrate that Euler’s and Froude’s model similitudes agree over wide ranges (for calculations, see Fig 8).
|
Fig.8 Simplified drawing of reservoir |
H.. Piezometric head B
Width of opening h Height of opening zo... Piezometric head top of theopening zu... Piezometric head at bottom of opening
|
In the case under study, an outlet opening 1m by 1m in cross section was investigated for a head varying between 1m and 10m. The flows according to Euler and according to Froude showed very good agreement down to small quotients of reservoir levels to height of the opening (for diagram, see Fig 9). The appendices show that Euler’s and Froude’s laws of similitude agree even for a quotient equal to 2, with a tolerance of 0.5%. Considering the velocity distribution in the outlet the quotient should be clearly higher (in our case the quotient is 20).
