Ferramenta para construção de modelos para análise isogeométrica de materiais compósitos

Abstract

The Finite Element Method (FEM) is currently the most widely used numerical method for structural analysis. The Isogeometric Analysis (IGA) is an alternative to FEM that uses for representation of the models the same functions used in software CAD as B-Splines and NURBS. In IGA, the same isoparametric formulation of the FEM is used, but the displacements are approximated using the functions that describe the model geometry. With this, the geometry approach error occurring in FEM is eliminated. Another advantage of IGA over FEM is ease of refinement. NURBS are described by a set of control points and parametric value (knot vector), providing representations of both standard models (quadrics, circles and conic sections) and free form using the same database. Because it is an even recent method, there are few programs using IGA, being more common academic software, such as FAST developed in the Laboratório de Mecânica Computacional e Visualização in the Universidade Federal do Ceará, the difficulty of using IGA in FAST is based on the pre-processing of data. In view of this, in this work, research of software with NURBS formulation was carried out in order to adopt one as a data pre-processing tool for IGA. Therefore, criteria of choice were taken such as cost, accessibility of NURBS data, and communication between the files generated by the program to choose graphical modeling software using NURBS and FAST. Rhinoceros, a software widely used for this purpose in the academy community, was adopted in this work to create models for linear analysis, stability, and vibrations analysis. Using Rhinoceros, a Python code was created capable of generating a file containing all NURBS data. Attributes such as loading, contour conditions, type of analysis, and material properties were associated through FAST_ModDef, other software developed in LMCV, generating the input file for analysis in FAST. The analyses were made for isotropic materials and composite materials, comparing the results obtained with the Literature and MEF, obtaining excellent results.