Author(s): Eszter Barta; Geza Hajnal; Gyongyi Karay; Vilmos Vasvari
Linked Author(s): Vilmos Vasvari
Keywords: Sand box model; Model laws; Hydraulic conductivity; Pumping test; Numerical modelling
Abstract: To determine the Darcy’s coefficient of permeability are several methods available. Empirical and deterministic calculation methods were developed of which applicability and accuracy depends on the available data and the type of investigated soil. Both field and laboratory investigations are common. In practice of civil engineering it is most essential task prior excavation to determine this soil physical parameter for planning of dewatering systems. Field investigations play central role also in the determination of recoverable water resources. In practice it is not common that all data required for the field investigation - usually pumping test - and its evaluation are available, the well design and the conditions of the measurement do not meet those assumed in the theory. Due to information of poor quality and anomalous conditions the calculated coefficient of permeability and the seepage hydraulic parameters can differ from the real values. The aims of the investigations were to conduct laboratory model tests in different soil types, also in their layered structure and by different design of the pumping well, to evaluate their results supported by numerical modelling and to come to conclusions which can be helpful in the areas mentioned above. In the course of the measurements size fraction and features of the pumping well were varied in order to achieve realistic field conditions. A laboratory model integrated also the field experiences was created. A cylindricallysymmetrical model with a ground plain of a quadrant, a radius of 1. 325 m and a height of 1. 0 m was used. By the numerical models both the field conditions and the experiments on the laboratory model could be simulated. The applicability limits of the most frequented calculation method for pumping test evaluation (Dupuit, Theis, Cooper & Jacob) were verified, analytical results were confronted with numerical ones. Two scales of models were analysed by computer: the first type had the scale 1: 16, which is the scale of the physical model and the second type had the scale 1: 1 modelled the real conditions. The two runs had the same results in three dimensional analyses, but the two dimensional analysis caused difference. The numerical models are suited to investigate soil parameters like in the laboratory but it hasmany advantages over the real laboratory tests. Moreover by means of the investigation’s results recommendations can be made for the layout of field tests (number of observation wells, distance of wells), for the type of the hydraulic test (conventional pumping test, single well test, slug test) and for the best applicable evaluation method.
Year: 2013