Modification of Cylindrical Load-Transfer Analytical Method by a Hybridized Automation Process

Abstract

There is a lot of research work on compression piles, but very little on tension piles, as commented by Chattopadhyay (1994), Patra and Pise (2003), Gaaver (2013), Lo et al. (2018) and Oh et al. (2019). The stress distribution in the soil mass surrounding an axially loaded tension pile is even not clearly understood, as Lo et al. (2018) discussed. This research has adopted an automation process to investigate the load-transfer factor in the cylindrical load-transfer analytical method in elastic and non-elastic soils for tension piles. Hybridized automation has been setup. It combines analytical and finite element methods to review and cover the deficiency, and modify the cylindrical load-transfer analytical method. The simulation of the load-displacement responses in tension piles becomes more realistic and efficient. Also, recent publications from Wang et al. (2012), Sheil and McCabe (2016), and Boonyatee and Lai (2017) have already demonstrated that using slipping analytical models can achieve more realistic and accurate results in compression piles. However, many researchers still adopt the simplified models by ignoring the pile-soil slipping behavior (the existing non-slipping analytical models), and no research has been conducted using slipping theory in tension piles. This research has been conducted to cover this deficiency in the analytical models of tension piles, as summarized in Lo et al. (2020). Furthermore, the existing load-transfer factor in the analytical method was evaluated only in elastic soils, which is a fixed value during axial loading. No further research detailly investigates its validation outside the elastic soils, and it is still a question to apply the factor outside these soils. In this research, a hybrid method and new equations have been developed to cover the deficiency and better understand how the factor will be used more appropriately. Additionally, automation processes save time, increase efficiency, lower cost, replace human labor with machine labor to minimize error. However, numerical methods are becoming more sophisticated and require considerable numerical expertise and computational resources, but too common to use. There is potentially a dangerous consequence (Mar, 2002). This research has developed a hybrid procedure for numerical results being crosschecked by the analytical method, similar to the current engineering practice of engaging a verifier to check the design work. It would be favorable to pile experts to carry out routine piling design or a quick initial estimation during the earlier stage of piling work. It sheds some light on the application of powerful and potential automation in research and engineering practice.