Banca de QUALIFICAÇÃO: LUCAS THIAGO GOMES SOARES DE ARAUJO

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
STUDENT : LUCAS THIAGO GOMES SOARES DE ARAUJO
DATE: 31/08/2026
TIME: 09:45
LOCAL: https://us02web.zoom.us/j/3010205143?pwd=VXVBTmhmV1k0dTZJclhvK3Evd1RPUT09
TITLE:

Monitoring and Prediction of the Structural Integrity of Offshore Wind Turbine Foundations: A Proposed Tool

KEY WORDS:

Offshore wind energy; System dynamics; Structural integrity; Fatigue and corrosion; Rio Grande do Norte.

PAGES: 139
BIG AREA: Engenharias
AREA: Engenharia de Produção
SUMMARY:

The expansion of offshore wind energy consolidates itself as a strategic pillar for the global energy transition and the achievement of power grid decarbonization goals. In the Brazilian context, the coast of Rio Grande do Norte stands out due to its high generation potential, driven by constant winds and shallow waters that favor the installation of fixed monopile structures. However, operation in an aggressive marine environment subjects these structures to complex degradation mechanisms, such as fatigue and corrosion, whose non-linear interactions challenge traditional methods of lifespan prediction. Given this complexity, this dissertation proposes a hybrid simulation architecture that couples Machine Learning with System Dynamics (SD), integrating three-dimensional physical accuracy with the systemic and temporal analysis of the asset. The overall objective of this work is to develop a monitoring and decision-support tool grounded in System Dynamics and powered by a surrogate Artificial Intelligence model to predict the structural integrity of offshore wind foundations in the Potiguar Equatorial Margin. Structured as a collection of five scientific papers, the research path encompasses a systematic review of recent literature, the analytical characterization of local hydrodynamic forces via the Morison Equation, the development of an AI model using Random Forest to instantaneously emulate Finite Element Analysis (FEA) mechanical responses, the formulation of a stock-and-flow model based on Sterman's iterative cycle, and, finally, the detailed explanation and modeling architecture of the integrated predictive monitoring system. The results reveal that the AI surrogate model achieves high predictive accuracy with low computational cost, enabling its real-time coupling. The systemic simulation evidences that the synergistic interaction between thickness loss due to corrosion and fatigue damage accumulation accelerates structural failure exponentially over a thirty-year horizon, eliminating the endurance limit of the steel. It is concluded that the developed hybrid model functions as a robust virtual laboratory, providing agile decision support in offshore engineering, mitigating the risks of catastrophic failures, and optimizing the planning of regional predictive maintenance policies.

COMMITTEE MEMBERS:
Presidente - 2456706 - MARIO ORESTES AGUIRRE GONZALEZ
Interna - ***.653.984-** - PRISCILA GONCALVES VASCONCELOS SAMPAIO - UFERSA
Externo à Instituição - HUMBERTO DIONISIO DE ANDRADE - UFERSA