Paper (rus)

Laboratory research of bedload discharge fluctuations

KONDRATYEV A.N.

BASIC EXPERIMENT DATA

Experiment was realized by the author under the direction of M.M. Gendelman in 1996 in the channel processes department of State hydrological institute in St.Petersburg, Russia.  Middle-bar type of channel process was reproduced in the experiment. Experiment was operated in the flume, width 2 m, length 50 m with the constant water discharge is equaled of 10 l/s.

At first, water poured in the basic tank by the pump with productivity 80 l/sec, and then water discharge in a flume was regulated by a valve on a head tank of the flume with a weir of a triangular profile, gabion and foam wave absorbers in a sump. The superfluous water discharge from the main tank was flown over at the top. It ensures a constancy of a head and, accordingly, water discharge, poured in the flume. The nil of the flume tank was determined by a volumetric method, and then the values of water discharges were calculated through a head on a weir according to the King formula. The flume end was equipped with an adjustable horizontal edge. 40 squared papers were glued together glasses sides every 1 meter. The zero points were determined using a horizontal surface of immovable water. Before the beginning of experiment the bottom was laid out from sand by an mean diameter of 0.33 mms and is ironed under a plane with slope 0.0022 manually by screw rack on two guides for glasses, which one were exhibited on the squared papers, glued on the flume glasses .The water filling of the flume on the laid out sand was made from below end of flume by small water discharge up to depth about 3 cm in a top of a flume.

Then monitoring measurements of depths on longitudinal section was conducted every 1 meter. The experiment was operated at constant flow of water, equal 10 l/sec. In the flume end closely to each other were established 8 bathometers for catching volume of bed load. The bathometers represent rectangular boxes 25 sm width. The upper edge of boxes was placed below basement of moved dune. Measurements of sand volume entered bathometers were made each 2 hours.

FLUCTUATION OF THE TOTAL BED LOAD DISCHARGE.

The experiment was prolonged 300 hours. In figure 1 the course of values bed load discharge are showed Temporal bed load irregularity is visible. This fact is known.In 1952 N.E. Kondratyev has formulated one of the basic concepts the channel processes theory. It is that "... in real channel flows all formations and all processes have discrete nature.... The discrete nature has also bed load transport" [1].

The fluctuation of the total bed load has essence relevant value for definition of indispensable accumulation period of the bathometer for obtaining representative value of the discharge bed load. Let's conduct conditional analogy. The fluctuation of the bed load discharge are similar to velocity pulsations of water flow. Only temporary scales of these fluctuations various. For obtaining water velocity we is maintained a whirly in a point 60-100 sec, and then we speak: “flow velocity there are a lot of meters per second”. Also for obtaining true average bed load discharge it is necessary to summarize values for hundreds hours, and then to receive the discharge in volumetric values per hour.

To evaluate attitude of these temporary scales, we compare time of replacement of all water in a flume (or on a unit of length) on new and propagation time by the microforms from a beginning of a flume up to the end (or unit of length). The water discharge is Q=10 l/s, width В = 2 m, depth H=0.03 m, length L=40 m. Water volume is peer 2.4 м³. Time of water in the flume is t=V/Q = 240 sec. Dune velocity Vd = 2 cm/hour, length of a flume L=40 m. A propagation time by the sand replacement of all flume t = L/ Vd = 2000 hours. Characteristics changes pace of a channel are more slowly pace of water replacement in 30000 times. If average value of flow rate of water we receive by average for 100 sec, that, guessing some analogy, bed load discharge it is necessary averaging for 3·10⁶ sec or approximately for 1000 hours!

The question are appears from the conclusions: How long time it is necessary to measure bed load discharge? Whether it is possible in general to measure a bed load discharge using bathometers?

CHANGE OF DISTRIBUTION OF A BED LOAD DISCHARGE ON WIDTH OF A FLUME.

The bathometers for measuring of the bed load discharge represent of eight closely put boxes of identical width. The obtained data, reduced in figure 2, allow observing change of the bed load discharge on width of a flume. About 100 hours the most part of bed load goes in any 3-4 adjacent boxes closer to a side. In another traps the bed load do not go at all or go in minimal quantities. Then approximately at 10 hours total quantity of bed load discharge is decreases and the bed load are arranged on width disorderly. Following hours almost all bed load discharge begin to go already in other 3-4 boxes for an inverse wall of a flume. Available no uniform distribution of the bed load transport on width of a flume. Non-uniformity is very large. For two hours in several boxes the bed load able not appears at all. While in any one box more than 50 % from the total bed load discharge for this period able is massed.

From here there are new question: How to select a place for the installation of bottom pan-type sampler on the river? Whether has sense a bed load measurement by bathometers in points? What in general represent the data of existing measurements?

Now it is possible to receive explanation, why any existing formula under the discharge of bed load does not respond the data of full-scale measurements. At first, in a nature never it is impossible by dot measurement to determine true value of the bed load discharge the integrating on width. Same it is possible to say about measurement of flow rates of water by a whirly. Really, thus too it is impossible to receive precise average value of flow rate of water, but the error will be in two, maximum three times. The estimation of average discharge of bed load on one measurement on width of the river threatens with an error in thousand times. Depending on a place of measuring it is possible to receive value and in hundreds time smaller, and in tens time large, than true one. And, secondly, as is described in the previous, the large enough time of average for obtaining true value is necessary. The precise measured values of the bed load discharge are not. Therefore there are no also valid formulas on calculation of the bed load discharge.

The unique report on continuous measurement of total bed load discharge on the river is an east Wind-Fork in USA [2]. The sampling of total bed load discharge implements the trap designed for continuous of a sampling on all width of a flow at low and average discharges of water and on separate segments of cross section at high flow of water. The total bed load entered the trap are forwarded on beach with the help of a driving belt and overturning bunkers, where they collect and are continuously weighed.

DEPENDENCE OF LARGE-SCALE CHANGES OF THE BED LOAD DISCHARGE ON WIDTH WITH MESOFORM’S TRANSIT.

For a problem of definition of the cause of formation channel bars and meander the prime value for us has dependence of distribution of the bed load discharge on width of a flow with transit mesoforms. In considered experiment was watched middle-bar type of channel process. This type of channel process is with no motion of sand in a forward and central part middle-bar. The bed load discharge there is equal to zero. The described scheme of the bed load transport by a cardinally differs from that, which one exists at side-bars.

At side-bars type of channel process, to the contrary, the most part of the bed load discharge goes on side-bar peripherals, and in the reach the transport is gentle. The qualitative difference of distribution of the bed load discharge on width of a channel is those at side-bar type of channel process and middle-bar.

 At middle-bar the maximum of bed load discharge goes on depression between middle-bar. At side-bar type of channel process, to the contrary, in reach the minimum of the bed load discharge is watched. This polarity in properties is conditioned to miscellaneous kinds of mesoform streamlining qualitatively by different velocity distribution of flow above middle-bar and side-bar. If an altitude of elevation small, the flow is easy flows through it and above such elevation there is an increase of velocity. If the altitude of elevation considerable, the flow can not skip through it, mesoform streamlines and above it there is velocity decreasing.

At middle-bar type of channel processes the flow streamlines middle-bar and the maximum both flow rates of water, and bed load discharge is necessary on reductions. At side-bar type of channel process, to the contrary, flow by oblate section flows through side-bar, the maximum of velocity is on side-bar that results in the fissile transport of bed load on such mesoform. Important qualitatively miscellaneous interplay of a flow and be load at side-bar and middle-bar. At middle-bar type the most discharge located in a middle-bar body participations actually do not receive, it is though superfluous for a flow and is streamlined by a flow. At side-bar type the flow bed load transport located on a side-bar surface.

If to see from the power point of view, at middle-bar the most of a flow does not suffice on the transport of all bed load offered by the river. At side-bar type of channel process the flow all quite suffice for the valuable fissile transport offered bed load. At side-bar type of channel process the energy of river has a lot of, it is not enough of bed load transport. At middle-bar type of channel process, to the contrary, the energy for the transport of bed load has not enough, the conveying capacity of a flow is rather small, it is a lot of bed load transport.

Middle-bar type of channel process is contrast of side-bar type.

REFERENCES

1. Kondratyev N.E. Popov. I.V., Snishenko B.F. Osnovy hydromorphologicheskoi teorii ruslovogo processa (Basic of hydromorphological theory of channel processes). – St. Petersburg: Gidrometeoizdat, 1982. – 272 с. (In Russian).

2. Lewin J., Bradley S.B., Macklin M.G. Historical valley alluviation in mid-Wales//J. Geol. - 1983. - Vol. 18, № 4. - p. 331-350

Fig. 1. Fluctuation of the total bed load discharge.

Fig. 2. A course of the bed load discharge measured each bathometer.

1-8 - number of boxes.