PREDICTION OF UNRAINED SHEAR STRENGTH AND CHARACTERISTICS OF CONSOLIDATION INDONESIAN CLAY SOIL USING MACHINE LEARNING

• Main Author: Maulana, Rizal
• Format: Theses
• Language: Indonesia
• Subjects: Teknik sipil
• Online Access: https://digilib.itb.ac.id/gdl/view/64143
• Institution: Institut Teknologi Bandung

The soil compression index (Cc) is a very important parameter to know the magnitude of soil decline, egitu also undrained shear strength (Su) to find out the carrying capacity of the soil with. Cc is obtained from the oedometer test and Su is obtained from triaxial, Direct Shear, and Unconfined Compressive Strength tests where the data is not yet available in the early stages of design so it is necessary to predict its value. This paper discusses the latest correlation model to predict the Cc and Su compression index of its properties index. The soil data used in the study came from various locations in Java, Sumatra, Kalimantan, and Papua. Many research reSults found a relationship between Cc clay soil and index properties Such as water content, liquid limit, and pore numbers while Su has a relationship with the SPT value or with the plasticity index. Cc used for this study has a strong correlation with its moisture content so it is used as a major predictor. Su used for this study has a strong correlation with N SPT so it is used as a major predictor. The analysis was conducted using simple regression and artificial neural network (ANN). As a result of the analysis, the moisture content can predict Cc as much as 73% and has a root mean square error of 0.12. The Cc value prediction equation from the ANN analysis results is recommended to be used at a moisture content of less than 60%. The SPT value can predict Su at most 69% and has a fairly large root mean square error. The Su value prediction equation from the ANN analysis results is recommended to be used on N60 less than 20.