2.4　Experimental study on transport rate of graded sediment⁽⁶⁾

Wang Shiqiang, Zhang Ren

Abstract: An investigation of the sediment transport rate of graded sand has been conducted in a 60m tilting flume in the Sediment Research Laboratory of Tsinghua University. The results of the experiments were used to verify the model which the authors proposed in 1989 for computing the transport rate of non-uniform sediment. In this model, mechanics of grain saltation, vertical distribution of velocity and sediment concentration and the interaction between coarse and fine grains at the river bed were considered. Good agreement is obtained in the comparison between the experimental data and the predictions of the model.

Procedure and results of experiment

The experiments were carried out in the Sediment Research Laboratory of Tsinghua University with a tilting flume 60m long and 1.2m wide. The dischargesused in experiment were up to 500L/s supplied and regulated by two large and one small pumps. Natural sand was used and its D₅₀ and D₆₅ were 0.078 and 0.086mm respectively with a non-uniformity coefficient σ(=0.5 (D₈₄/D₅₀+D₅₀/D₁₆)) of 1.44. The gradation curve is shown in Fig. 1.

Fig. 1　size distribution of model sand

At the beginning of each test, a sand layer 15cm thick was placed on the flume bed. The discharge Q and the water depth h were considered as independent variables in the experiments. Following self-adjustment of the bed configuration, the equilibrium energy slope S and the sediment concentration C were formed dependently. The duration of each test depended mainly on the flow intensity. It ranged from 4 to 20 hours in which the equilibrium state of fluvial adjustment could be reached. The discharge was measured by an electro-magnetic current meter which was calibrated against a weir. The longitudinal water surface profile was measured at 36 locations along a working the section of 50m using an automatic water surface follower. Due to the existence of a long working section, the accuracy in determining the energy slope was much greater than that which can be achieved in a short flume. Both sides of the flume are made of glass. This enabled the flow depth to be measured at 36 locations in the 50m working section on both sides of the flume using a ruler. The sediment concentration in the flow was determined using a siphon tube at nine points in a cross-section at the entry of the flume where the flow is very turbulent and the distribution of sediment concentration is quite uniform. The volume of each sediment sample was approximately of 1000mL.

The average water surface slope was calculated from the measured water level using a least-squares linear regression. The average flow depth was calculated by averaging the 36 measured depths.

Twenty-seven sets of experiment on the transport capacity of sediment in the flow were carried out and the main results of the experiment are summarised in Table 1. The ranges of discharge, water depth, energy slope and sediment concentration varied from 32 to 410 1/s, 0.100 to 0.365m, 0.554 to 3.054×10^-3 and 4.71 to 188.95kg/m³ respectively. According to the size distribution of bed material, 0.025mm can be roughly regarded as the critical diameter between wash load and bed material load in flow. In Table 1, C is the measured total sediment concentration, T is the temperature, S is the energy slope, h is the water depth and Q is the discharge.

Table 1　Results of transport rate experiment

The composition of bed material changes to some extent when the flow intensity changed. But, for simplification we used a constant bed material size distribution as shown in Table 2 in the calculation thereafter.

Table 2　Size distribution of bed material