TY - JOUR

T1 - Optimization of fiber distribution in fiber reinforced composite by using NURBS functions

AU - Ghasemi, Hamid

AU - Brighenti, Roberto

AU - Zhuang, Xiaoying

AU - Muthu, Jacob

AU - Rabczuk, Timon

N1 - Funding information: This work was supported partially by Marie Curie Actions under the Grant IRSES-MULTIFRAC and German federal ministry of education and research under the Grant BMBF SUA 10/042. Nachwuchsförderprogramm of Ernst Abbe foundation, the National Basic Research Program of China (973 Program: 2011CB013800), Program for Changjiang Scholars and Innovative Research Team in University (PCSIRT, IRT1029) and Pujiang Program (12PJ1409100) and the research support provided by the Italian Ministry for University and Technological and Scientific Research (MIUR) is also acknowledged.

PY - 2013/12/24

Y1 - 2013/12/24

N2 - This research deals with the optimization of short fibers distribution in continuum structures made of Fiber Reinforced Composite (FRC) by adopting an efficient gradient based optimization approach. Motivated by lack of non-heuristic and mesh independent optimization algorithms to obtain the optimum distribution of short fibers through a design domain, Non-Uniform Rational B-spline (NURBS) basis functions have been implemented to define continuous and smooth mesh independent fiber distribution function as well as domain discretization. Thanks to higher order (here quadratic) NURBS basis functions along with their compact support, a drastic reduction in computational time has been obtained by increasing mesh size while the accuracy of the model is maintained. Moreover combination of NURBS with sensitivity based optimization method allows a fast convergence to optimum fiber distribution layout. Minimization of elastic strain energy and maximization of fundamental frequency have been considered as objective functions for static and free vibration problems, respectively; to get the maximum fiber exploitation in the structural element. Nodal volume fraction of fiber was defined as the optimization design variable while a homogenization approach based on the random orientation of short fibers in the matrix has been adopted. Some numerical examples related to the structural response under static loading as well as the free vibration behavior are finally conducted to demonstrate the capability and reliability of the model.

AB - This research deals with the optimization of short fibers distribution in continuum structures made of Fiber Reinforced Composite (FRC) by adopting an efficient gradient based optimization approach. Motivated by lack of non-heuristic and mesh independent optimization algorithms to obtain the optimum distribution of short fibers through a design domain, Non-Uniform Rational B-spline (NURBS) basis functions have been implemented to define continuous and smooth mesh independent fiber distribution function as well as domain discretization. Thanks to higher order (here quadratic) NURBS basis functions along with their compact support, a drastic reduction in computational time has been obtained by increasing mesh size while the accuracy of the model is maintained. Moreover combination of NURBS with sensitivity based optimization method allows a fast convergence to optimum fiber distribution layout. Minimization of elastic strain energy and maximization of fundamental frequency have been considered as objective functions for static and free vibration problems, respectively; to get the maximum fiber exploitation in the structural element. Nodal volume fraction of fiber was defined as the optimization design variable while a homogenization approach based on the random orientation of short fibers in the matrix has been adopted. Some numerical examples related to the structural response under static loading as well as the free vibration behavior are finally conducted to demonstrate the capability and reliability of the model.

KW - Fiber distribution optimization

KW - Fiber Reinforced Composite (FRC)

KW - NURBS

KW - Objective function

KW - Optimization

UR - http://www.scopus.com/inward/record.url?scp=84890815467&partnerID=8YFLogxK

U2 - 10.1016/j.commatsci.2013.11.032

DO - 10.1016/j.commatsci.2013.11.032

M3 - Article

AN - SCOPUS:84890815467

VL - 83

SP - 463

EP - 473

JO - Computational Materials Science

JF - Computational Materials Science

SN - 0927-0256

ER -