|
Computational & Technology Resources
an online resource for computational,
engineering & technology publications |
|
Civil-Comp Conferences
ISSN 2753-3239 CCC: 14
PROCEEDINGS OF THE SIXTEENTH INTERNATIONAL CONFERENCE ON COMPUTATIONAL STRUCTURES TECHNOLOGY Edited by: P. Iványi, J. Kruis and B.H.V. Topping
Paper 6.2
An Efficient Multiscale Topology Optimization Framework Using HiCon-FEM V. Yanes1, N.-H. Kim2, M.A. Sanz1 and F. Montáns1
1Escuela Técnica Superior de Ingeniería Aeronáutica y del Espacio, Universidad Politécnica de Madrid, Spain
Full Bibliographic Reference for this paper
V. Yanes, N.-H. Kim, M.A. Sanz, F. Montáns, "An Efficient Multiscale Topology Optimization Framework Using HiCon-FEM", in P. Iványi, J. Kruis, B.H.V. Topping, (Editors), "Proceedings of the Sixteenth International Conference on
Computational Structures Technology", Civil-Comp Press, Edinburgh, UK,
Online volume: CCC 14, Paper 6.2, 2026, doi:10.4203/ccc.14.6.2
Keywords: HiCon-FEM, SIMP, topology optimization, hierarchical condensation, multiscale optimization, Gaussian filter.
Abstract
This paper presents a multiscale topology optimization framework that combines the
classical Solid Isotropic Material with Penalization (SIMP) method with the Hierar-
chical Condensation Finite Element Method (HiCon-FEM). The aim is to preserve
the simplicity and robustness of density-based topology optimization while reducing
the computational cost associated with the repeated solution of large fi nite element
systems. The formulation is solved with an Optimality Criteria update scheme, and
a Gaussian convolution-based fi lter is employed as a scalable regularization strategy
for large meshes. The main contribution is the use of HiCon-FEM as a drop-in re-
placement for the standard fi nite element solver. Numerical results for the MBB beam
benchmark show speedups of up to 40.0× for a 900 × 300 discretization and up to
35.0× for a 3000 × 2000 discretization while preserving the main topological features
of the classical SIMP solutions.
download the full-text of this paper (PDF, 2623 Kb)
go to the previous paper |
|