Numerical investigation on pile group efficiency embedded in soft clay Abdelkrim Ferchat, Mohamed Nabil Houhou and Sadok Benmebarek MN2I2S Lab, Department of Civil Engineering and Hydraulic, Mohamed Khider University of Biskra, Biskra, Algeria Abstract Purpose – This paper aims to investigate the pile group efficiency based on the load-settlement response in soft clay conditions, considering several pile configurations using a variable number of piles and pile spacing. The overall objective of the present paper is to provide further insight into the mechanical response of the pile group and aim at helping the engineers in taking a logical path in an iterative design process for pile group efficiency. Design/methodology/approach – To investigate the pile group efficiency, three-dimensional (3D) numerical simulations were performed using the finite-difference code FLAC3D. Findings – The obtained numerical results are validated by comparing them to those of similar subgrade structure and in comparable geological conditions provided within the literature. The results indicated that although the bearing capacity of the pile group increases with increasing number of piles, the efficiency of the pile group is very important for a small number of piles. However, increasing of pile spacing has a positive effect on pile group efficiency depending on piles number and settlement level. The pertinence of the 3D numerical results of efficiency coefficient is judged by comparison with those obtained from the most popular formulas available in the literature. Originality/value – A predicted model is also proposed which is validated with the obtained numerical results to a better goodness of fit. Keywords Numerical analysis, Settlement, Efficiency coefficient, FLAC3D, Pile group, Soft clay Paper type Research paper Introduction Shallow foundations usually respond to the design criteria of geotechnical foundations in terms of bearing capacity and settlement. However, in cases of inappropriate soil conditions and high applied load, the need to transfer the applied load to deeper and hard soil is indispensable. The foundation solution is therefore to build deep foundations. Deep foundations are widely used in tall buildings, bridges, towers, offshore and other special structures. This type of foundation is designed to transmit the applied forces to the surrounding soil at a significant depth which usually exceeds ten diameters. The deep foundations are often group of piles where, in practice, piles are always driven or bored by group. Because of the effect of the interaction between piles, the behavior of a group of piles may be different from that of individual pile and the limit vertical load of a group of n piles (Q GP ) may be less than n times the limit load of an individual pile (Q SP ) working under the same conditions. The group effect is evaluated using the efficiency coefficient of the bearing capacity of pile group (C g ) which is defined as the ratio of Q GP to n times Q SP . In the past century, some investigations have been carried out to define the value of the efficiency coefficient C g such as the research works of Feld (1943), Whitaker (1957), Saffery and Tate (1961), De Mello (1969), Barden and Monckton (1970), Brand et al. (1972), O’Neill et al. (1982) and Briaud et al. (1989). Most experimental evaluations of C g are applied depending only on soil conditions and the pile installation method. In conditions of loose to medium dense sand, C g = 1 for driven piles and for bored friction piles C g tends to lower values. However, in clay soils, C g is often lower than unity (de Sanctis and Mandolini, 2006). Nevertheless, the efficiency coefficient also depends on the number of piles and pile spacing (Cooke, 1986; Frank, 1999). Empirical formulas have also been developed for the evaluation of the efficiency factor of the pile group such as the Converse–Labarre method (Bolin, 1941) or the Los Angeles group action method (Das, 2015). Unlike the other methods, these empirical formulas only consider the plan geometry of the foundation. Despite the research efforts carried out to date to assess the pile group efficiency and their encouraging results, discrepancies are still observed during comparison between the results of the formulas already developed and field or laboratory test measurements (Helmy, 2002; Park and Lee, 2015). Numerical methods such as finite difference, finite element and boundary element method, have been widely developed in the past two decades because they are less costly and may be The current issue and full text archive of this journal is available on Emerald Insight at: https://www.emerald.com/insight/1708-5284.htm World Journal of Engineering © Emerald Publishing Limited [ISSN 1708-5284] [DOI 10.1108/WJE-04-2020-0112] The authors express their gratitude to the Directorate General for Scientific Research and Technological Development of Algeria. Received 6 April 2020 Revised 5 September 2020 Accepted 8 November 2020