Determinação da taxa de infiltração da água através de sensores termopares e de capacitância (FDR) em um solo areno-siltoso remoldado em laboratório

Abstract

Infiltration measurement systems are not standardized. Traditional equipment is strongly influenced by factors that cannot be controlled, requires mobilization of the construction site, and has limited use in the laboratory. On the other hand, modern measuring equipment has high costs and usually provides indirect data that needs to be interpreted to obtain the infiltration rate. In this context, the present study aims to determine the water infiltration rate in a sandy-silty soil remolded in the laboratory through thermocouple and capacitance sensors (FDR) compatible with Arduino. The developed equipment will improve the infiltration measurement to obtain it in a direct, fast, and economical way. For this, a soil composed of 70% of medium sand and 30% of silty soil of the Guabirotuba Formation, both dry, where the fines were excluded to avoid expansion, was developed. Granulometry test, real grain density (Gs), Atterberg limits, X-ray Fluorescence Spectrometry (FRX), X-ray Diffractometry (DRX), Scanning Electron Microscopy (MEV), and Proctor compaction at standard, intermediate, and modified energy for the materials were carried out. The infiltration tests were performed in three replicates for each humidity point of the Proctor curve with the sandy-silty soil at the modified energy. The cylinders used for molding the samples were made of acrylic, thereby allowing the visualization of the wetting front. The infiltration inside the sample was determined by equipment composed of a microcontroller (Arduino), an internet connection device (ESP 8266), and temperature (thermocouple) and humidity (FDR) sensors embedded in a fiberglass board. With the equipment data, the infiltration per time curves were drawn for the sandy-silty soil created, and the infiltration rate of this soil at different molding moistures was fitted to mathematical equations. From the results, it is possible to highlight that the time for water to infiltrate in the driest soil was approximately twice as short as in the wettest soil, taking 7.1 minutes and 14.1 minutes, respectively. However, after a certain humidity where the soil was very wet, its ability to retain water was compromised, causing the infiltration to become faster. This behavior can be observed in the samples molded with 13.8% and 10.3% of initial moisture. The wetter samples had a higher infiltration rate (81.16 mm.h-1) than the samples with 10.3% (71.31 mm.h-1). The empirical equations developed to determine the infiltration rate were all powers with the base as the time it took to the water to infiltrate. When comparing them with the other empirical methods for determining the infiltration rate already consolidated in the literature (Kostiakov, Kostiakov-Lewis and Horton), the Horton model was the one that best fitted the tests performed. Finally, the equipment was validated through the capacitance sensor (FDR) and visual tests of the wetting front.