Paper publicado no periódico Composite Structures, por: Prof. Dr. Vergílio T. S. Del Claro, Prof. Dr. Aldemir Ap. Cavalini Jr., Prof. Dr. Ilmar F. Santos, Prof. Dr. Valder Steffen Jr.
Disponível em: https://doi.org/10.1016/j.compstruct.2022.116030
Abstract
"Recent times have seen a great interest on environmental issues and efficient, sustainable systems. This interest has required the employment of advanced composites for a myriad of industrial machines and innovative equipments. Among these applications, Flywheel Energy Storage Systems - FESS - represent a group of machines that are being re-invented through this process. Modeling composite flywheels has proven to be a complex task, which current Finite Element models fail to fulfill in a number of design contexts. This demand to model complex composite geometries and systems induced the proposition of new methods, aiming to capture the various physical effects existing in the problem. In the present contribution, the authors consider that it is viable to model the dynamic behavior of a Flywheel Energy Storage System via an adapted Carrera Unified Formulation, both in terms of accuracy and computational cost, for practical applications. The present work presents and explores a Carrera Unified Formulation model with extended capabilities dedicated to rotordynamics applications. The differences from standard Finite Elements models are presented, evidencing advantages and drawbacks of the proposed methodology over more traditional approaches. A case study is then presented, modeled, and the results are compared with those stemming from established formulations.".
From the presented results, it can be concluded that the present rotor tailored CUF formulation is capable of representing a complex rotor dynamic behavior, going as far as modeling the flywheel as a shaft element. Not only that, but also its connection to the metallic shaft and associated dynamic interactions. A noteworthy point is that this model does not use optimization to adjust mechanical properties to those stemming from experimentation. Despite of that, the model produced quite close matching results. This raises two relevant points to observe regarding the formulation conveyed: (a) the numerical implementation of rotor-tailored CUF is computationally light, making optimization procedures viable; (b) the formulation represents the physical phenomena correctly and with satisfactory detail, being able to obtain even physically irrelevant terms that would have been otherwise neglected. Regarding (b), while it might be unnecessary to consider all the physical terms all the time, they do not increase the execution time significantly and the disturbances can be smoothed by performing tracking or amplitude based analyses. Besides, it should be pointed out that while some terms may seem irrelevant at a given operation condition they may as well be dominant for another condition or even interact with other modes in unexpected ways, whose simpler models would fail to predict.
Fig.1: FESS experimental test-rig;
Fig.2: FEM design and proposed mesh;
Fig.3: Experimental modal analysis performed, with the first two modes presented;
Fig.4: Orbits of the FESS open-end at different rotating speeds, comparing the operation in horizontal and vertical condition;
Fig.5: Experimental Bode diagram for the FESS test-rig;
Fig.6: Comparison between numerical run-up results for the 1st and 2nd order UF models. It is clear that the 1st order model does not predicts the correct behavior, while the 2nd order one does so to a satisfactory extent;
Fig.7: Numerical Campbell diagram produced for the FESS open-end;
Fig.8: Numerical Campbell diagram for the FESS, comparing the tracked and untracked calculation procedures, making evident the necessiti for a tracked Campbell calculation.
Comments