Helmut Drobir 1,
Volker Kienberger 2 and Johannes Seyerl 1
1
Institute of
Hydraulic Engineering, Vienna University of Technology,
A
-1040 Wien, Karlsplatz 13/222, Austria/Europe
Tel:
+43-1-58801-22230, Fax: +43-1-504-59-28
E-mail:
helmut.drobir@tuwien.ac.at
2
Va Tech Hydro
Gmbh & Co,
A
–4031 Linz, Lunzerstraße 78, P.O.Box 28, Austria/Europe
Tel:+43-70-6987-8817,
Fax:+43-70-6980-8826
E-mail:
Volker.Kienberger@vatech-hydro.at
Abstract:
Investigations were performed using a hydraulic model
of tunnel-type high-head leaf gates to study downpull effects. The model
investigations were carried out for particular headwater and tailwater
conditions, for different deep gate slots and for the most common top seal and
skinplate arrangement which is with seal and skinplate on the downstream side.
According to the tailwater conditions free-surface flow and submerged flow have
occurred. For the free-surface flow, the resulting downward hydraulic forces of
the hydraulic model test, the so-called downpull, are compared with results of
calculation according to the method of Naudascher [2].
Keywords: downpull, leaf gate, tunnel-type high-head gate, gate slots, free-surface flow, submerged flow, hydraulic model
High-head gates operate under heads larger than their height. The most widely used high-head gates for flow regulation or emergency closure of large conduits are leaf gates. Leaf gates can operate in a gate well located within a conduit transition and offer many advantages in construction and maintanance. From all types of high-head gates, leaf gates cause the greatest problems in connection with hydrodynamic forces. The pressure along the bottom surface of the gate is reduced during operation because of high efflux velocities. The accompanying downward hydrodynamic forces, the so-called downpull, are often greater than the dead weight of the gate.
In most cases the arrangement for tunnel-type gates is to have top seal and skinplate on the downstream side. A reduction in downpull occurs, if the seals of a tunnel-type gates are arranged on the upstream side. But this advantage of reduction downpull is negated by the extensive vortex action in the gate slots. The downward hydraulic force (downpull) controls the dimension of the hoist mechanism. The prediction of this downpull has been, and still is, of major concern in design and it affects the safety of the entire project. But a moderate downpull is usually welcomed to secure safe gate closure under emergency conditions.
For high-head gates two states of flow occur: The free-surface flow in which the space downstream of the gate is filled with air; and the submerged flow in which the whole tunnel is filled with water. The performed investigations were carried out for both states of flow. The free-surface flow depends on the efficency of the ventilation through an air inlet and on the dimensions of the cross section of the tunnel on the downstream side of the gate. For this free-surface flow, the downpull of the hydraulic model test was compared with results of calculation according to the method of Naudascher [2]. On the upstream side of the gate the tunnel is always submerged.
Fig. 1 Definition sketch of a tunnel-type gate for submerged flow
In the sketch above, H is the total
head, Q is the total inflow, Q’ is the rate of discharge under the gate,
Q’’ is the rate of discharge over the gate, y0 is the height of
the pressure tunnel, y is the gate opening, yc is the jet
contraction, h is the piezometric head (water level) in the air vent,
the gate bottom inclination and the
other symbols are explained in figure 1.
The prototype of the gate was 5.2 m high, 5.6 m wide and 1.25 m deep. For the investigations on hydraulic forces effecting on tunnel-type high-head leaf gates, a wet shaft arrangement was modeled on a scale of 1: 25 in the Hydraulic Laboratory of the Institute of Hydraulic Engineering in Vienna. The wet shaft arrangement was composed of the storage reservoir formed by a steel container, the pressure tunnel and the gate well, both made from plexiglas, so that the flow processes could be well observed. The cross section of the pressure tunnel was rectangular. The leaf gate was a fixed roller gate. The body of the model gate was made from sheet brass and the rollers consisted of stainless steel. The hoist mechanism for opening or closing the gate was of a screwed spindle moved by an electric step motor. A steel bar of 5mm in diameter connected the gate with the screwed spindel.
The lifting and lowering velocity of the gate model was 1 cm/min. This speed of gate operation dt/dy was very small compared to the flow velocity vc near the gate. So it was a common assumption that it was permissible to disregard the unsteady effect caused by the the gate movement on the downpull. The unsteadiness parameter is given by the formular
(1)
The lifting forces, acting in the direction of the hoist mechanism, are composed of three forces: the dead weight minus buoyancy, the friction force against moving direction, and the downpull. When the model gate was moved downwards as well as upwards, the lifting forces were measured by a pressure gauge. Hereby the friction forces could be eleminated. Since the dead weight minus buoyancy of the gate is known, the downpull is calculable.
Fig. 2 Model of the wet shaft arrangement
The Froude law of
similarity was used for modelling hydraulic forces and for studying downpull
effects. The model tests were carried out for free-surface flow and submerged
flow. For free surface-flow, the contraction of the jet released under the gate
depends on the Froude number of the jetflow. The Froude number of the jet flow
(1) is determined by the formular:
(2)
in which vc denotes the velocity of the contracted jet, Cc is the concentration coefficient and y is the gate opening. This Froude number Fr was restricted to small values. Therefore it can be neglected when one deals with high-head gates and Fr > 4. In the model tests the Froude number was always greater than 4, depending on the piezometric heads in the upstream steel container and depending on the ratio gate opening from 0.35 to 0.45, where maximum downpull appeared.
There are also model-scale effects to viscosity.
Adapted from Naudascher [3] the model-scale effects to viscosity were negligible
for Reynolds numbers:
(3)
According to Naudascher [3], the Reynolds numbers
(4)
for submerged flow
without cavitation, was large enough to be representative of prototype
conditions. In this formula, vc is the velocity of the contracted
jet, d is the thickness of the gate and
is the kinematic fluid viscosity.
In all phases of the model tests the Reynolds number exceeded Re = 1.65·10 5.
Recent studies by Naudascher have shown, that the approach - flow conditions may have great effects on the downpull characteristics. This would explain the different results of the maximum downpull determined by a semi-empircal approach after Naudascher [2] and the maximum downpull found by hydraulic model tests carried out by Kienberger [1].
The test results on maximum downpull for two gate-slot-ratios and two gap-ratios are presented in Figure 3 to 6 and compared with the results of the method of Naudascher. The gate-slot-ratio is
(5)
and the gap-ratio is defined as
(6)
The symbols of this formula are explained in Figure 1.
The maximum downpull occures in the ratio y/y0 of the gate opening from 0.35 to 0.45. In Figure 3 to 6 the non dimensional downpull is defined by the ratio F/FH.
F is the maximum downpull and FH is the pressure force as a result of the total head H of water in the reservoir.
(7)
H/y0 is the ratio of the total head to the height of the pressure tunnel. The model tests were carried out for 4 different heads of water levels in the reservoir (H/y0 = 4, 8, 12 and 15). In relation to the prototype the ratio of 15 was equivalent to 75 m total head in a reservoir.
|
|
|
|
|
|
Figures 3 – 6: These figures show the maximum downpull ratio F/FH and the total head ratio H/y0 for free-surface flow and submerged flow, for different depth of gate slots (in non dimensional form = CB) and for different sizes of the gaps (in non dimensional form = Ca2) between the skinplate of the gate and the recess shown in Figure 1.
The maximum downpull is obtained for ……
(1) free-surface flow by means of semi-empirical approach by Naudascher [2].
(2) free-surface flow from model tests (water as model fluid) by Kienberger.
(3) submerged flow h/H = 0.2 from hydraulic model tests.
(4) submerged flow h/H = 0.5 from hydraulic model tests.
(5) submerged flow h/H = 0.8 from hydraulic model tests.
The hydraulic model tests for submerged flow were carried out by Seyerl [4].
The results presented in Figures 3 –
6 are for gates with a closed upstream face and a defined form of the gate lip.
The geometrical parameters of the gate lip are the gate bottom inclination
and the gate thickness d. Both are
shown in Figure 1. Additional model tests were carried out for gates with
identical form of the gate lip, but the upstream face of the gate was open. The
results of this investigations are mentioned in the conclusion of this paper.
The downpull on the hydraulic model of a tunnel-type high head leaf gate, with seal and skinplate arranged on the downstream side, was determined for free-surface flow and submerged flow. The results of the hydraulic model test were compared with results determined by a semi-empirical approach according to Naudascher.
The maximum of downpull occured in the ratio y/y0 of the gate opening from 0.35 to 0.45.
Regarding the effect of flow passing over the gate, an inlarging of the gap a2 between the skinplate of the gate and the recess of the gate well reduce the downpull. It may thus be advantageous to increase the rate of overflow by changing the spacing.
The downpull was also reduced by gate slots. The
larger the gate slot relative to the gate, the greater the reduction; and the
reduction was maximum for an intermediate range of y/y0.
It is important whether the upstream face of the gate is open or closed. If the face is composed of girders open to the approach flow, the lower-most girder has an effect on the streamlines. The streamline will be directed slightly away from the gate bottom and this will reduce the downpull.
The hydrodynamic forces (downpull) generated from hydraulic model tests were lower than the computed downpull based on the method developed by Naudascher. The results according to the semi-empirical method of Naudascher are obviously on the safe side.
References
[1]
Kienberger V.: Hydrodynamische Kräfte an Tiefschützen, Ph.D. Thesis, Vienna
University of Technology, 1999.
[2] Naudascher E.: Hydrodynamic Forces, IAHR/AIRH Hydraulic Structures Design Manual 3, A.A. Balkema, Rotterdam, 1991.
[3] Naudascher E.: Scale Effects in Gate Model tests. In: Symposium on Scale Effects in Modelling Hydraulic Structures, H. Kobus (ed.). Institut für Wasserbau, Universität Stuttgart, Germany, 1984.
[4] Seyerl J.: Hydrodynamische Kräfte an Tiefschützen in Abhängigkeit vom Einstaugrad, Master Thesis, Vienna University of Technology, 2000.
Fig. 5 CB = 0.25