CFD Analysis of an Articulated-Body Wave Energy Converter: Comparison with Experimental Data and Frequency-Domain Modeling Predictions

Abstract

Developing sustainable energy systems requires continued advancement in renewable energy technologies. Among these, wave energy is considered a particularly promis ing resource due to its substantial theoretical potential. This thesis investigates an articulated-body wave energy converter using a computational fluid dynamics (CFD) model developed in STAR-CCM+. To evaluate the predictive capability of the numerical model, previously conducted model tests are replicated under regular wave conditions. Wave periods ranging from 0.7 to 2.2 seconds are considered, with the hinge-angle amplitude as the primary response quantity. This response is particularly important, as it is directly related to the relative motion available for energy extraction. The simulation results are compared with experimental measurements and results from a previously developed frequency-domain model. The numerical setup includes wave generation, multibody dynamics, and key subsystems such as the mooring and Power Take-Off (PTO) systems. Two PTO formulations are implemented: a linear formulation and a fitted nonlinear formulation based on the measured torque-angular velocity relationship from the model tests. Overall, the CFD models show close agreement with the experimental data across the complete range of analyzed cases. They capture the experimentally observed peak hinge-angle response and reduce overprediction near the peak compared with the previous frequency-domain model. All CFD models predict the peak response at the same wave period as the experiments and within 10% of the experimental peak amplitude, whereas the frequency-domain model overpredicts the peak by more than 30%. Comparisons between the two PTO representations further indicate that this subsystem significantly influences the predicted response across the investigated cases. Among the investigated approaches, the CFD model with the fitted nonlinear PTO formulation provides the best agreement with experimental measurements, with a global error of approximately 10%. However, the comparison also shows that the improved agreement cannot be attributed solely to the hydrodynamic model, as the PTO formulation significantly influences the predicted hinge-angle response. These findings indicate that reliable prediction of the WEC response requires not only accurate hydrodynamic modeling but also representative subsystem input data, particularly for the PTO system.