[Invited speech]Reducing Installation Uncertainty of Plate Anchors Using a Symbolic Regression-Based Macro-element Model

Reducing Installation Uncertainty of Plate Anchors Using a Symbolic Regression-Based Macro-element Model
ID:117 Submission ID:122 View Protection:ATTENDEE Updated Time:2026-07-30 15:35:00 Hits:1 Invited speech

Start Time:2026-08-10 16:40 (Asia/Hong_Kong)

Duration:15min

Session:[S4] Session 4 Marine Geo-disaster and Geo-environment » [S4] Session 4 Day 2

No files

Abstract
Suction-embedded plate anchors (SEPLAs) are promising anchoring solutions for deep-water offshore structures. However, their installation involves large rotation and translation under eccentric mooring-chain loading, making the final anchor embedment and orientation difficult to predict. Since these installation characteristics directly affect anchor capacity, efficient trajectory prediction is essential. Large-deformation finite-element (LDFE) methods can capture anchor–soil interaction but are computationally expensive, whereas macro-element (ME) models provide an efficient alternative. Their predictive accuracy, however, depends strongly on the assumed yield surface and flow rule describing anchor failure behavior.
This study investigates the limitations of existing ME formulations and develops an improved ME framework for plate anchor installation. A database of three-dimensional finite-element simulations is used to examine commonly adopted yield functions under combined vertical force, horizontal force, and moment loading. The results show that existing formulations may not accurately represent the failure envelope and plastic flow behavior under complex loading conditions, resulting in uncertainties in predicted anchor trajectories.
A symbolic regression-based approach is therefore proposed to identify an improved mathematical representation of the anchor yield surface. By systematically exploring physically meaningful interaction terms, the proposed method establishes a yield function that better captures anchor failure behavior while satisfying the requirements of an elastoplastic ME formulation. The improved yield function is incorporated into an elastoplastic ME model for efficient trajectory prediction. Validation against LDFE simulations demonstrates that the proposed model accurately reproduces anchor trajectories and force responses with substantially reduced computational cost.
The developed framework provides a systematic approach for reducing installation uncertainty of plate anchors by improving the constitutive description of anchor–soil interaction and offers an efficient tool for offshore anchor design.
Keywords
plate anchor,macro-element,offshore foundations,soil-structure interaction,keying
Speaker
Maozhu PENG
The Hong Kong Polytechnic University

Submission Author
Maozhu PENG The Hong Kong Polytechnic University
Zhen-Yu YIN Hong Kong Polytechnic University
Comment submit
Verification code Change another
All comments
Log in Sign up Registration Submisison