Deblurring structural edges in variable thickness topology optimization via density-gradient-informed projection

Area of Science:

• Engineering
• Computational Mechanics
• Materials Science

Background:

• Variable thickness topology optimization (VTTO) is crucial for designing high-stiffness sheet structures.
• Existing VTTO methods face challenges with manufacturing difficulties due to low-thickness regions and blurred structural edges from regularization filters.

Purpose of the Study:

• To address the challenges of low-thickness regions and blurred edges in VTTO.
• To introduce a novel density-gradient-informed (DGI) projection for enhanced edge definition and manufacturability.

Main Methods:

• A combined approach using SIMP-based penalization and an updated projection method to suppress low-thickness domains.
• Development and application of the density-gradient-informed (DGI) projection, utilizing local density gradients to sharpen edges.

Main Results:

• The proposed methods effectively suppress nearly all low-thickness regions.
• The DGI projection successfully deblurs structural edges, restoring a distinct solid-void transition without altering internal structure.
• Edge definition improvements were achieved with negligible impact on structural compliance.

Conclusions:

• The DGI projection is an effective, non-invasive regularization tool for VTTO.
• This method enhances the manufacturability and precision of VTTO designs.
• The approach significantly improves edge definition in topology optimization while maintaining structural performance.