《蝶阀动水力矩数值分析》由会员分享，可在线阅读，更多相关《蝶阀动水力矩数值分析（6页珍藏版）》请在金锄头文库上搜索。
1、A Numerical Prediction of the Hydrodynamic Torque acting on a Safety Butterfly Valve in a Hydro-Electric Power Scheme A. D. Henderson, J. E. Sargison, G. J. Walker and J. Haynes School of Engineering University of Tasmania Hobart AUSTRALIA alan.hendersonutas.edu.au http:/www.eng.utas.edu.au Abstract
2、: - A numerical study of the flow through a safety butterfly valve used in a hydro-electric power scheme to stop water supply to a downstream penstock is reported. Computational fluid dynamics applied in a quasi-steady manner is used to predict the variation in hydrodynamic torque coefficient with o
3、pening angle during a constant head test. Factors influencing these results, such as Reynolds number and unsteady flow effects, are found to be significant. The predicted results are compared with field measurements of the full-size valve. Issues associated with applying the numerical results to pre
4、dict valve characteristics at higher Reynolds numbers are discussed. Further computational and experimental studies are recommended. Key-Words: - Numerical, Safety, Butterfly valve, Torque, Hydro-electric power 1 Introduction Hydro-electric power schemes require safety valves to stop water flow from
5、 the reservoir to the turbine. Turbines that operate under significant head and have a long penstock usually need a valve near the upstream reservoir to isolate flow to the penstock. Butterfly valves are often chosen for this purpose since they have a simple mechanical construction, fast closing tim
6、e, and more importantly give a low head loss when fully open (see 1). The size of these hill-top valves will depend on the size of the penstock section in which they are housed. Larger valves may have diameters up to 5 m. The valve investigated in this study is installed in a hydro-electric power ge
7、nerating scheme located in central Tasmania, Australia. The valve has a diameter D = 3.048 m and a leaf of convex cross- section. The valve closing mechanism is similar to that described by Ellis and Mualla 2. The valve is normally in a fully open position. When the valve is triggered by either an a
8、bnormally high flow rate, or a remote controlled signal, a locking mechanism disengages. The valve is then subjected to a large out of balance moment imposed by weights that act to close the valve. The closing motion of the valve is regulated using two large oil-filled dashpots. Oil is forced from t
9、he dashpot chambers through a small orifice, which gives a smooth and near constant valve closing rate. During a valve closure test, the dashpot pressure, valve position, upstream static head and downstream static head are recorded. The torque acting on the valve is then estimated from the measured
10、dashpot pressure and the mechanical arrangement of the valve. The torque acting on a closing valve may be resolved into several components (see American Water Works Association (AWWA) 3): bearing torque (Tb), torque imposed by an offset centre of gravity of the valve (Tcg), hydrodynamic torque (Td),
11、 packing torque (Tp), and torque due to hydrostatic pressure (Th). The sign convention used in this study is for torque to be positive when acting in the closing direction. Components Tb and Tp always act in the opposite direction to the valve closing direction. Components Td and Tcg may act in eith
12、er direction, depending on geometry 2, 3. It is common to express the hydrodynamic torque Td in the form of a dimensionless torque coefficient such as Ct = Td / (D3 P) where P represents the static pressure differential across the valve. AWWA 3 suggests for model testing, that the downstream pressur
13、e should be measured at least 10D downstream of the valve to allow for sufficient pressure recovery and the upstream pressure should be measured at least 2D upstream. A common method of determining the valve torque characteristic of a butterfly valve is by a constant head test as described in AWWA 3
14、. In WSEAS TRANSACTIONS on FLUID MECHANICSA. D. Henderson, J. E. Sargison, G. J. Walker and J. HaynesISSN: 1790-5087218Issue 3, Volume 3, July 2008that test, the valve is positioned in a long horizontal section of pipe that is fed by a constant head source. Measurements of torque and head loss are t
15、hen made at various valve angles. The actual head-flow characteristic will differ from that in such simple model tests because there will be greater operating head, additional head loss components due to penstock friction, bends and transitions; and also a head drop across the turbine. In an actual
16、test, the turbine head reduces as the valve closes, leading to very low pressure behind the valve. To prevent buckling collapse of the penstock, anti-vacuum valves located a short distance downstream from the valve admit air into the penstock tunnel when the pressure drops below atmospheric. This is also known to reduce the severity of cavitation 3, 4. Owing to these factors, the maxim
