Hidrológiai Közlöny 1968 (48. évfolyam)

8. szám - Varrók Endre: A nyomásvonal alakulása töltésezett folyók mellett

346 Hidrológiai Közlöny 1968. 8. sz. Varrók E.: A nyomásvonal alakulása 0,80 0,80 0,80 0,80 r 1 o,eo m Fig. 3. Section of experimentál setup F, A — overflow weirs, l — bitumen layer, 2 — air vents, :t - porous concrete partition, •/ — concrete layer, .5 — pressure-gauge outiets, fí — riprap of stone cliips 3. ábra. A kísérleti berendezés metszetei Jelfllésck: F és A vízszintszabályozók, I bitumenréteg, 2 légtelenítő csövek, 3 porózus beton fal, é betonréteg, S piezométeres megcsapolások, 0 zúzottkő ment in both, covered and uncovered aquifers: the only diference was in the structure of the constant eoefficient. Checking of the tlieory could most simply have been carried out by analyzing field observations. However, the number ofground water level observations available was insufficient to do so. A further difficulty was that the full thickness of the aquifer was not disclosed in most of the cases, i. e. an important boundary eondition was unknown. For this reason, the validity of the theoretical solutions was checked by model experiments. The results of these will be described in the chapter to follow. Simultaneously the evaluation of further field observations has been conimenced. This work has not yet been brought to a t'lose. The experimentál setup The porous médium involved in the experiment was a fine sand with a skewness U = 2,25 and an effective grain size dn = 0,22 min as determined by Kozeny's procedure. The sand was filled into a basin 5 m long and 0,3 m wide (Fig. 3). After building in the partitions of porous concrete and filling the two end compartments with a riprap of stone cliips, the basin was filled with water to a depth of 0,4 m and the sand was filled in under water. The pore volume of the layer thus förmed was n = 0,40; its coefficient of permeability, k = 9,52-10-'

A cm/sec (at 21°C water temperature). The sand was covered with 3 cm of concrete and the latter with 5 cm of bitumen. In order to prevent the softening of the bitumen and to ensure an appropriate confining pressure, water was poured on top of the bitumen in a depth of 0,5 m. There was no communication between this water and the water in the sand. Air trapped in the sandy aquifer could escape through closeable vents brought out through the concrete cover. To permit the observation of pressure changes, the basin was provided with 11 pressure gauge outiets connected to glass piezometer tubes of 4,5 mm i. d. Pressure changes were recorded by photographing the tubes once per minute. This technique and the fourfold repetition of each flood wave permitted to record 1000 to 1300 pressure data per experiment, depending on the length of the flood wave. The end compartments were connected to overflow weirs. The overfiow level of both was set at 0,445 m heiglit above the upper surface of the sand layer, whereas one of them could be moved up and down, from the above-mentioned level as its median position. Because of the above-described structure of the experimentál setup, no free water table could develop in the aquifer, the water was under pressure throughout. The boundary conditions resulting from this setup were the following: there was no seepage towards the imperrneable side walls, base or cover. In- and outflow were possible at both ends of the sand prism at the same velocity in the totál endsection. The piezometric pressure expressed in height of water column was h = constant at one end surface and a function of time: h=h

0 (0; t) at the other. The long axis of the prism was chosen xaxis: the piezometric pressure was constant over any cross section perpendicular to the x-axis; all pressure changes were in the x direction. The boundary eondition h = h

0 (0; t), i. e. the function pressure change vs. time could be arbitrarily chosen. The infinite rangé of possibilities was, however, limited by the requirements that the function should be easy to handlé mathematically and easy to produce by mechanical means. With due attention especially to the second requirement, we have chosen the periodic function shown in Fig. 4. The period of the function is T; its rangé (double amplitude) is H; it is described mathematically by the Fourier series A

o(0; t)-H 8 k=0 1)" (2k + l)

2 sin[(2&+l)cof| (4) The value of the function varies over the rangé from +H/2 to —H\2: the extremes occur at the instants t = nT/4 (where re is an oddpositiveinteger). In (4), the cyele frequency co = 2n characterizes the periodic function, whereas the phase angle cot = ot