Evaluation of the group effect’s simulation on the bearing capacity of bored piles in granular soil, using load transfer functions

• Autor(a) principal: Cerqueira Júnior, Edno
• Data de Publicação: 2022
• Outros Autores: Moura, Alfran Sampaio
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Federal do Ceará (UFC)
• Texto Completo: http://www.repositorio.ufc.br/handle/riufc/70948
• topic: Load transfer functions; Numerical analysis; Prediction of settlements; Prediction of bearing capacities; Piled rafts

Load transfer functions, such as the t-z and q-z curves, combined with numerical methods implemented by computational programs are nowadays widely used in geotechnical design of deep foundations. It is known that in this type of design, piles are more often employed in groups instead of isolated. Within this context, the purpose of this research was to estimate the bearing capacity of groups of bored piles, installed in granular soil, using t-z and q-z curves suggested by Reese and O’Neill (1988). The geotechnical design program RS Pile was used in the simulation of an isolated pile’s load–displacement curve. The result was used along with group settlement factors (ξ) in order to predict the load-displacements curves of several pile groups, the graphics obtained were used to estimate the bearing capacities of the groups. Lastly, all the predictions were compared with experimental data obtained from load tests carried out in foundations installed in the Experimental Field of Geotechnics and Foundations of the Federal University of Ceará (CEGEF-UFC). For the isolated piles, the difference between the simulated and experimental values of bearing capacities ranged from -11% to 5% and the settlement predictions were also accurate. For the pile groups, regarding the settlement prediction, the compared curves showed agreement for the design load for 6 out of the 7 groups analyzed. Regarding the bearing capacity, the difference between the predicted to the experimental values ranged from -8% to 20%. Thus, the employed method was able to produce accurate predictions.