Published Date
15 November 2016, Vol.156:10–19, doi:10.1016/j.compstruct.2016.05.058
70th Anniversary of Professor J. N. Reddy
Author
C.Y. Kiyono a,b,,,
S.L. Vatanabe a
E.C.N. Silva a
J.N. Reddy b
Topology optimization
Stress limit
Composite structures
Multi-p-norm
SIMP
Load independent
For further details log on website :
http://www.sciencedirect.com/science/article/pii/S0950061814013312
15 November 2016, Vol.156:10–19, doi:10.1016/j.compstruct.2016.05.058
70th Anniversary of Professor J. N. Reddy
Author
aDepartment of Mechatronics and Mechanical Systems Engineering, Politechnique School of University of São Paulo, Avenida Professor Mello Moraes, 2231, 05508-900 São Paulo, SP, Brazil
bDepartment of Mechanical Engineering, Texas A&M University, College Station, TX 77843-3123, USA
Received 1 January 2016. Revised 15 May 2016. Accepted 18 May 2016. Available online 20 May 2016.
Abstract
Although the stress-based topology optimization problem has been extensively studied for continuum structures, it is still an open problem and there is still room for improvements. This work proposes a comprehensive approach for dealing with stresses in topology optimization problems. The SIMP method is used to distribute material along the domain. For limiting the stress, a multi-p-norm formulation is proposed to deal with the local nature of stress and to avoid stress concentration. This function considers many values of p coefficients at the same time while other formulations adopt a specific value for p defined for subregions. As a consequence this formulation can avoid the stress concentrations without being necessary to define sub-regions. In addition, the proposed formulation is load independent because the multi-p-norm is used as the objective function. A SIMP-like formulation is used to address the stress singularity phenomenon and the heaviside projection is used to avoid mesh dependency, checkerboarding, and to control the minimum length-scale. A proper continuation scheme is proposed to all penalization coefficients in order to achieve black-and-white solutions. The optimization problem is solved by using GCMMA. Numerical examples for homogeneous and composite structures are presented to illustrate the proposed formulation.
Keywords
- ⁎ Corresponding author at: Department of Mechanical Engineering, Texas A&M University, College Station, TX 77843-3123, USA.
For further details log on website :
http://www.sciencedirect.com/science/article/pii/S0950061814013312
No comments:
Post a Comment