A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Prediction of Bearing Capacity of Stone Columns Underlying a Soil–Cement Bed Using Support Vector Regression Approach
Stone columns are normally used to improve the bearing capacity and reduce the compressibility of soft clayey soil. But in the case of very soft clay, stone columns cannot take much load over it, it bulges and hence failure occurs. In those cases, a hard layer of soil–cement bed can be placed over the stone columns to increase the bearing capacity of stone columns. Use of a soil–cement bed (SCB) over stone columns reduces the stress intensity over the columns and increases the load-carrying capacity of stone columns floating in very soft clay. With this concept, a set of laboratory experiments were performed by varying the length (l) of stone columns, spacing (s) between the stone columns, and thickness (t) of the soil–cement bed. From the experimental investigation, the bearing capacity of soft soil was observed to be improved by 8.5 times with the use of a group of three stone columns underlying the soil–cement bed (SCB). As there is no well-established equation for the determination of the bearing capacity of stone columns underlying SCB till today, therefore, the aim of the present study is to develop an empirical model for the estimation of bearing capacity (qu) of stone columns underlying SCB and to form an empirical equation to get qu from the results of the empirical model. For this, a total of 180 data were collected from the experimental investigation and used for the training and testing of numerical models. Two numerical models have been developed by using two types of artificial intelligence (AI) method namely support vector regression (SVR) and adaptive neuro-fuzzy inference system (ANFIS). The former method gives a better result with a coefficient of regression value of R = 0.99951, hence it is used to obtain the precise empirical formula of the determination of qu. The formula has been validated with some experimental data which were not used in developing the numerical models. The results of this study showed that SVR could be a powerful alternative physical tool for the prediction of bearing capacity of the stone columns underlying SCB.
Prediction of Bearing Capacity of Stone Columns Underlying a Soil–Cement Bed Using Support Vector Regression Approach
Stone columns are normally used to improve the bearing capacity and reduce the compressibility of soft clayey soil. But in the case of very soft clay, stone columns cannot take much load over it, it bulges and hence failure occurs. In those cases, a hard layer of soil–cement bed can be placed over the stone columns to increase the bearing capacity of stone columns. Use of a soil–cement bed (SCB) over stone columns reduces the stress intensity over the columns and increases the load-carrying capacity of stone columns floating in very soft clay. With this concept, a set of laboratory experiments were performed by varying the length (l) of stone columns, spacing (s) between the stone columns, and thickness (t) of the soil–cement bed. From the experimental investigation, the bearing capacity of soft soil was observed to be improved by 8.5 times with the use of a group of three stone columns underlying the soil–cement bed (SCB). As there is no well-established equation for the determination of the bearing capacity of stone columns underlying SCB till today, therefore, the aim of the present study is to develop an empirical model for the estimation of bearing capacity (qu) of stone columns underlying SCB and to form an empirical equation to get qu from the results of the empirical model. For this, a total of 180 data were collected from the experimental investigation and used for the training and testing of numerical models. Two numerical models have been developed by using two types of artificial intelligence (AI) method namely support vector regression (SVR) and adaptive neuro-fuzzy inference system (ANFIS). The former method gives a better result with a coefficient of regression value of R = 0.99951, hence it is used to obtain the precise empirical formula of the determination of qu. The formula has been validated with some experimental data which were not used in developing the numerical models. The results of this study showed that SVR could be a powerful alternative physical tool for the prediction of bearing capacity of the stone columns underlying SCB.
Prediction of Bearing Capacity of Stone Columns Underlying a Soil–Cement Bed Using Support Vector Regression Approach
Indian Geotech J
Das, Manita (author) / Dey, Ashim Kanti (author)
Indian Geotechnical Journal ; 55 ; 1168-1183
2025-04-01
16 pages
Article (Journal)
Electronic Resource
English
Bearing capacity of geosynthetic encased stone columns
| Tema Archive | 2013
Bearing capacity of group of stone columns
| British Library Conference Proceedings | 2006