Predicting failure envelopes of helical anchors under combined VHM loading using a regression-enhanced knowledge graph attention network
Article
| Article Title | Predicting failure envelopes of helical anchors under combined VHM loading using a regression-enhanced knowledge graph attention network |
|---|---|
| ERA Journal ID | 4710 |
| Article Category | Article |
| Authors | Vichai, Katavut, Shiau, Jim, Tran, Duy Tan, Khajehzadeh, Mohammad, Keawsawasvong, Suraparb and Jamsawang, Pitthaya |
| Journal Title | Ocean Engineering |
| Journal Citation | 341 (Part 3) |
| Article Number | 122660 |
| Number of Pages | 39 |
| Year | 2025 |
| Publisher | Elsevier |
| Place of Publication | United Kingdom |
| ISSN | 0029-8018 |
| 1873-5258 | |
| Digital Object Identifier (DOI) | https://doi.org/10.1016/j.oceaneng.2025.122660 |
| Web Address (URL) | https://www.sciencedirect.com/science/article/pii/S0029801825023431 |
| Abstract | This study presents a regression-adapted Knowledge Graph Attention Network (KGAT) integrated with Water Cycle-Moth Flame Optimization (WCMFO) to predict the three-dimensional failure envelopes of helical anchors embedded in anisotropic clay under combined vertical-horizontal-moment (V-H-M) loading. A comprehensive dataset of dimensionless load responses, i.e., normalized vertical, horizontal, and moment capacities (V/Asu, H/Asᵤ, M/ADsᵤ) is generated using Finite Element Limit Analysis (FELA). Key input parameters include the helix diameter ratio (Dh/Ds), embedment ratio (L/Ds), strength anisotropy ratio (re), and load inclination angle (β), all of which influence the failure envelope topology. A hybrid KGAT-WCMFO model is developed to evaluate the FELA results, achieving high predictive accuracy with R2 values exceeding 0.97 for both H/Asᵤ and M/ADsᵤ. The predicted failure envelopes closely match those from FELA and effectively capture the nonlinear interactions among soil strength, anchor geometry, and load direction. These results highlight the model's potential as a robust surrogate tool for rapid and accurate geotechnical assessments under complex loading conditions. |
| Keywords | Helical anchors; Failure envelope; Knowledge graph attention network (KGAT); Water cycle-moth flame optimization (WCMFO); Surrogate modeling; Finite element limit analysis (FELA); Anisotropic clay |
| Contains Sensitive Content | Does not contain sensitive content |
| ANZSRC Field of Research 2020 | 400502. Civil geotechnical engineering |
| Public Notes | Files associated with this item cannot be displayed due to copyright restrictions. |
| Byline Affiliations | Thammasat University, Thailand |
| School of Engineering | |
| Islamic Azad University, Iran | |
| King Mongkut’s University of Technology North Bangkok, Thailand |
https://research.usq.edu.au/item/100w66/predicting-failure-envelopes-of-helical-anchors-under-combined-vhm-loading-using-a-regression-enhanced-knowledge-graph-attention-network
5
total views1
total downloads5
views this month1
downloads this month