Evaluation of Bearing Capacity of Clay Reinforced with Deep Mixing Columns and Geotextile Caps by Large-Scale Experiments & Artificial Neural Networks

Document Type : Original Article

Authors

1 PhD Student, Department of Geotechnical Engineering, Faculty of Civil Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran(farzadpourebrahim317@gmail.com).

2 Assistant Professor, Department of Civil Engineering, Yasuj Branch, Islamic Azad University, Yasuj, Iran

10.22034/road.2024.437750.2247

Abstract

Soil Deep mixing columns (DSM) is one of the methods to improve the resistance and bearing capacity of soil for the development of urban areas. The purpose of this research was to evaluate the carrying capacity of clay reinforced with deep mixing columns and geotextile caps with large-scale experiments and artificial neural networks. . First, the basic technical characteristics of the soil bed were obtained by preliminary laboratory tests, such as granulation type, natural percentage, Etterberg limits, pH, and uniaxial compressive strength. Then, determination of moisture content percentage and optimum chemical additive amount of soil was done by standard compaction tests and uniaxial compressive strength to apply in the construction of deep soil mixing columns in situations without and with geotextile cap. In the second stage of the laboratory tests, load-settlement diagrams were created for each scenario with suitable construction for the laboratory model, different configurations of cement columns and without conditions and with geotextile caps on the cement columns. Laboratory studies showed that the presence of cement columns can increase the bearing capacity of the foundation up to 18 times compared to their absence. The maximum vertical load applied at a constant deformation of 30 mm was equal to 18.37 and 24.59 kN in the cases without and with the geotextile cap, respectively, which showed a 33% increase in the amount of applied load. The artificial neural network for all four separate forms showed an acceptable level of accuracy with a correlation value of 0.94.

Keywords