@article{Domphoeun2026, 
author = {Rithy Domphoeun and Duy Tan Tran and Raksiri Sukkarak and Suraparb Keawsawasvong},
title = {Intelligent stability prediction of 3D contiguous I-shaped pile walls via finite element limit analysis and neuro-fuzzy optimization},
year = {2026},
journal = {Journal of Intelligent Construction},
keywords = {stability number, contiguous pile wall, I-shaped pile, finite elements limit analysis, adaptive neuro-fuzzy inference-genetic algorithm},
url = {https://www.sciopen.com/article/10.26599/JIC.2026.9180138},
doi = {10.26599/JIC.2026.9180138},
abstract = {This study presents a stability analysis on three-dimensional (3D) contiguous I-shaped pile walls with c-ϕ backfill soil by using the three-dimensional finite element limit analysis (3D FELA) method. The stability number ( γcrH/C) of the 3D contiguous I-shaped pile walls was computed based on 3D FELA with many factors, including soil friction angle (ϕ), wall–soil interface roughness (m), height-to-flange width of ratio (H/B), pile spacing-to-flange width ratio (S/B), and length-to-width or embedment depth ratio (D/B). The stability number of the contiguous I-shaped pile wall increased continuously with increasing ϕ, m, and D/B, whereas it decreased with increasing S/B. Moreover, a hybrid machine learning method, adaptive neuro-fuzzy inference-genetic algorithm (ANFIS-GA), was used to model all the datasets. The results indicated that ANFIS-GA performed very well in predicting stability number with coefficients of determination (R2) of 0.985 and 0.972 for the training and testing datasets, respectively.}
}