Session: 04-01-03 Flexible Pipes III

Submission Number: 181980

Numerical-Experimental Calibration of the Chaboche-Voce Model for Tensile-Armor Steels 

Residual stresses accumulated during cold forming, winding, and spooling of tensile-armor wires strongly influence the susceptibility of flexible-riser steels to CO₂-induced stress-corrosion cracking (SCC-CO₂). Because those stresses arise from repeated plastic straining in manufacture, reliable assessment of riser integrity requires constitutive models that can reproduce cyclic hardening, mean-stress relaxation, and ratcheting under the loading amplitudes of interest.

This study outlines a numerical–experimental procedure for calibrating the Chaboche–Voce model for a high-strength carbon steel used in tensile-armor layers. Experimental data were obtained from monotonic tension tests and strain-controlled cyclic tests performed on representative specimens. Multiple reversed cycles provided stabilized hysteresis loops exhibiting pronounced Bauschinger reversal, transient hardening, and mean-stress relaxation. The monotonic results supplied the initial yield stress, delimiting the parameter search space.

A representative volume element was analyzed with a three-component Chaboche–Voce formulation. The model parameters were identified through a genetic algorithm that minimized a hybrid objective function that combined the discrete Fréchet distance with root-mean-square errors at characteristic stress points. The optimization algorithm was applied iteratively until convergence.

Three model variants—three back-stress components, four back-stress components, and four components with static recovery—were compared. The three-term variant reproduced the cyclic response as accurately as the more complex alternatives while reducing computation time, and was selected for further analysis. Cross-validation, in which parameters calibrated on one specimen were applied to others, produced low hybrid correlation errors. Simulated hysteresis envelopes matched the experimental shape, amplitude, and mean-stress evolution without artificial ratcheting, indicating that the calibrated model captures the mechanisms governing steels susceptible to SCC-CO₂.

The proposed workflow integrates targeted laboratory testing, reduced-order finite-element simulation, and evolutionary optimization to generate consistent constitutive parameters suitable for process-induced residual-stress predictions and large-scale riser assessments. The framework can be extended to other alloys, alternative loading paths, or different optimization schemes, providing a systematic approach for offshore structural-integrity evaluations.

Presenting Author: Luis Chambizea Subsea Technology Laboratory

Presenting Author Biography: Luis Chambizea is a doctoral researcher at the Laboratory of Subsea Technology (LTS) at the Federal University of Rio de Janeiro (UFRJ). He holds a bachelor’s degree in Naval Engineering and a master’s in Oceanic Engineering from UFRJ. His research focuses on the cyclic plastic behavior of steels and its application to subsea structures, combining experimental testing of flexible pipes with advanced constitutive modeling and parameter optimization using genetic algorithms. His work aims to improve the predictive capabilities of material models for offshore applications.

Authors:

Luis Chambizea Subsea Technology Laboratory
Marcelo Igor Lourenço De Souza Subsea Technology Laboratory
Neilon De Souza Da Silva Petrobras