A novel digital unit cell library generation framework for topology optimization of multi-morphology lattice structures

Liu, Jinlong, Zou, Zhiqiang, Gao, Kang, Yang, Jie, He, Siyuan and Wu, Zhangming ORCID: https://orcid.org/0000-0001-7100-3282 2025. A novel digital unit cell library generation framework for topology optimization of multi-morphology lattice structures. Composite Structures 354 , 118824. 10.1016/j.compstruct.2024.118824 Item availability restricted.

PDF - Accepted Post-Print Version Restricted to Repository staff only until 20 December 2025 due to copyright restrictions. Download (7MB)

Abstract

Although single-unit cell lattice structures are commonly used in engineering, multi-morphology composite lattice structures offer enhanced mechanical properties and diverse functionalities by tailoring their microstructures. This study presents a novel framework for generating a digital unit cell library to optimize the design of multi-morphology lattice structures. The framework involves creating the library using modular encoding and an adjacency matrix while addressing connectivity constraints. A voxel model is employed to streamline the homogenization process of the various unit cells in the library. This homogenization dataset trains a Radial Basis Function Neural Network (RBFNN) to evaluate the elasticity tensor of the unit cells. The compliance of the multi-morphology lattice structure is minimized by identifying optimal unit cell volume fractions and types through topology optimization and 0-1 integer programming. The former utilizes sensitivity analysis via RBFNN to determine the optimal volume fractions, while the latter focuses on minimizing the element strain energy and considers constraints that satisfy the optimal volume fractions of unit cells. The effectiveness and feasibility of this method are demonstrated through three benchmark numerical examples. Experimental results from additive manufacturing samples show that the proposed multi-morphology lattice structures achieve a 44.24% increase in initial stiffness and a 47.70% increase in ultimate strength compared to single-unit cell lattice structures.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	Engineering
Subjects:	T Technology > TA Engineering (General). Civil engineering (General)
Publisher:	Elsevier
ISSN:	0263-8223
Funders:	EPSRC
Date of First Compliant Deposit:	7 January 2025
Date of Acceptance:	19 December 2024
Last Modified:	21 Jan 2025 12:30
URI:	https://orca.cardiff.ac.uk/id/eprint/175087

Actions (repository staff only)

Edit Item