<p>This paper focuses on the implementation of the nodal-based Floating Frame of Reference Formulation with model-order reduction. The paper introduces two variations of the nodal-based Floating Frame of Reference Formulation that neglect certain rigid-body degrees of freedom. The variations are demonstrated with simulations of a two-stroke marine engine crankshaft, discretized with finite element and significantly reduced using distributed coupling master/slave relations and the Component Mode Synthesis reduction technique. The master nodes introduce rotational degrees of freedom in the reduced-order model, and this paper makes an original contribution by modifying the nodal-based Floating Frame of Reference Formulation to account for moments applied to the master nodes’ rotational degrees of freedom. Simulations in the time domain compare the variations with benchmark results from a commercial multibody software. The results demonstrate the influence of the rigid-body degrees of freedom on the accuracy for the chosen application, and a fifteenfold computational efficiency gain can be achieved with a MATLAB implementation of the nodal-based Floating Frame of Reference Formulation, when compared with the commercial multibody software.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Nodal-based floating frame formulations for efficient marine engine crankshaft simulation

  • Jacob Østerby Rasmussen,
  • Andreas Zwölfer,
  • Ilmar Santos

摘要

This paper focuses on the implementation of the nodal-based Floating Frame of Reference Formulation with model-order reduction. The paper introduces two variations of the nodal-based Floating Frame of Reference Formulation that neglect certain rigid-body degrees of freedom. The variations are demonstrated with simulations of a two-stroke marine engine crankshaft, discretized with finite element and significantly reduced using distributed coupling master/slave relations and the Component Mode Synthesis reduction technique. The master nodes introduce rotational degrees of freedom in the reduced-order model, and this paper makes an original contribution by modifying the nodal-based Floating Frame of Reference Formulation to account for moments applied to the master nodes’ rotational degrees of freedom. Simulations in the time domain compare the variations with benchmark results from a commercial multibody software. The results demonstrate the influence of the rigid-body degrees of freedom on the accuracy for the chosen application, and a fifteenfold computational efficiency gain can be achieved with a MATLAB implementation of the nodal-based Floating Frame of Reference Formulation, when compared with the commercial multibody software.