Design-oriented predictive models for tubular T-joints based on nonlinear finite element and shakedown analyses
Journal article, 2026

Thin-walled tubular joints are fundamental load-carrying components in offshore and marine structures, where structural integrity under extreme and cyclic loading conditions is a key design concern. This study develops an integrated framework to predict the ultimate limit state and elastic shakedown limit of tubular T-joints subjected to axial compression. A comprehensive parametric finite element analysis was performed, and empirical design-oriented formulas were systematically derived through automated regression methods. The ultimate limit state was evaluated using the Riks method,yielding six predictive models; the best-performing model showed excellent agreement with numerical results (R2 = 0.9935, COV = 0.0340) and outperformed existing design standards, reducing mean absolute percentage errors from 23.81 % (API) and 30.68 % (CIDECT) to 15.59 %. The shakedown limit was estimated using the linear matching method, and six additional formulas were proposed.The most accurate model achieved R2 = 0.9979 and COV = 0.0190. Validation against incremental cyclic elastoplastic analyses confirmed the predictive capability, with 121 parametric cases correctly classified and the best formula reaching 97.66 % accuracy. The proposed formulas provide a practical design tool for thin-walled tubular joints, enabling safer and more efficient offshore structuralapplications.

Riks method

Linear matching method

Empirical formula

Tubular T-joints

Elastic shakedown

Ultimate limit state

Author

Yeong Je Kim

Seoul National University

Nak-Kyun Cho

Seoul National University of Science and Technology (SeoulTech)

Jonas Ringsberg

Chalmers, Mechanics and Maritime Sciences (M2), Marine Technology

Do Kyun Kim

Seoul National University

Ocean Engineering

0029-8018 (ISSN)

Vol. 350 124300

Driving Forces

Sustainable development

Areas of Advance

Production

Materials Science

Roots

Basic sciences

Subject Categories (SSIF 2025)

Materials Engineering

Applied Mechanics

DOI

10.1016/j.oceaneng.2026.124300

More information

Latest update

2/23/2026