The effect of ballasted column installation is commonly simulated using two main numerical approaches: (i) a mechanistic representation based on an expansive cavity, and (ii) the Wish-in-Place approach combined with an improved initial stress state (WIP + K0). The expansion cell method has been widely regarded as a robust, computationally efficient technique for simulating the effects of dry-stone column installation, whereas the WIP + K0 method is commonly used for wet-stone column installation. However, this method constitutes a simplified procedure presenting several technical weaknesses. In this study, a three-dimensional numerical simulation of an in-situ load–unload test on a stone column is conducted to investigate installation-induced effects. For this purpose, two numerical installation procedures are considered under the Plaxis 3D software environment. Firstly, the applicability of the conventional expansion cell method for simulating wet-stone column installation is investigated, taking into consideration the effect of the installation energy. Secondly, a novel stage-by-stage construction procedure based on a sequential installation of the column. In this methodology, the installation energy is simulated as a vertical imposed displacement applied to the top of each column sequence, which constitutes the principal novelty of this research. The vertical imposed displacement is chosen according to the literature review, indicating that, for a wet-installation procedure, installation energy does not involve a lateral pressure due to the presence of the installation fluid immediately. To validate the numerical results, a full-scale stone column load-unload test carried out at the Port of Algiers is simulated by applying a series of surface loads in accordance with the in-situ test load sequences, aiming to assess the effects of the installation process and compaction intensity. Finally, a comparison between the numerical and experimental load–unload curves shows that the proposed modelling approach reproduces the in-situ response of the ballasted column. The best results are obtained with an installation energy of 10 cm, with a medium error below 2%, demonstrating its efficiency for wet column installation. In contrast, the expansion cell method results in a significant underestimation of settlement, by as much as 36%, for load levels ranging from 75 to 125% of the column bearing capacity. At the final loading stage (150% of the column’s ultimate load), the numerical predictions are in good agreement with the experimental measurements. These results indicate that the cell expansion procedure can be efficient for simulating wet-column installation and predicting the ultimate capacity of stone columns, while the new step-by-step method is promising. However, this funding needs to be confirmed through additional simulations of in situ stone column loading tests.

3D-simulation of an isolated stone column-loading test: effect of the installation process

期刊论文
发表日期：

2026年4月13日

查看全文

Mohamed Chikhaoui,
Lynda Djerbal

摘要

The effect of ballasted column installation is commonly simulated using two main numerical approaches: (i) a mechanistic representation based on an expansive cavity, and (ii) the Wish-in-Place approach combined with an improved initial stress state (WIP + K₀). The expansion cell method has been widely regarded as a robust, computationally efficient technique for simulating the effects of dry-stone column installation, whereas the WIP + K₀ method is commonly used for wet-stone column installation. However, this method constitutes a simplified procedure presenting several technical weaknesses. In this study, a three-dimensional numerical simulation of an in-situ load–unload test on a stone column is conducted to investigate installation-induced effects. For this purpose, two numerical installation procedures are considered under the Plaxis 3D software environment. Firstly, the applicability of the conventional expansion cell method for simulating wet-stone column installation is investigated, taking into consideration the effect of the installation energy. Secondly, a novel stage-by-stage construction procedure based on a sequential installation of the column. In this methodology, the installation energy is simulated as a vertical imposed displacement applied to the top of each column sequence, which constitutes the principal novelty of this research. The vertical imposed displacement is chosen according to the literature review, indicating that, for a wet-installation procedure, installation energy does not involve a lateral pressure due to the presence of the installation fluid immediately. To validate the numerical results, a full-scale stone column load-unload test carried out at the Port of Algiers is simulated by applying a series of surface loads in accordance with the in-situ test load sequences, aiming to assess the effects of the installation process and compaction intensity. Finally, a comparison between the numerical and experimental load–unload curves shows that the proposed modelling approach reproduces the in-situ response of the ballasted column. The best results are obtained with an installation energy of 10 cm, with a medium error below 2%, demonstrating its efficiency for wet column installation. In contrast, the expansion cell method results in a significant underestimation of settlement, by as much as 36%, for load levels ranging from 75 to 125% of the column bearing capacity. At the final loading stage (150% of the column’s ultimate load), the numerical predictions are in good agreement with the experimental measurements. These results indicate that the cell expansion procedure can be efficient for simulating wet-column installation and predicting the ultimate capacity of stone columns, while the new step-by-step method is promising. However, this funding needs to be confirmed through additional simulations of in situ stone column loading tests.