Voxel-Based Geometry Reconstruction for Repairing and Remanufacturing of Metallic Components via Additive Manufacturing


In the component repair process using additive manufacturing (AM) technique, reconstructing the repair volume is essential to generate the repair tool path and guide the AM system to deposit correct geometry on the worn parts. In this study, a novel repair volume reconstruction methodology based on voxel modeling was introduced. At first, the stereolithography (STL) models of the nominal and damaged components were acquired either from computer-aided design (CAD) modeling or through robot-aided 3D scanning. The pre-machining approach was introduced to guarantee the defective zone is accessible to repair tools. The damaged models were aligned with the nominal models by best-fitting their common convex-hull centroids. After that, these STL models were converted to voxel models using the Marching Cube algorithm. The accuracy of the transferred voxel models was investigated by comparing them with the exact geometry. Boolean operations algorithm based on constructive solid geometry was proposed to extract the repair volume by comparing the nominal and damaged voxel models. The Boolean operations can output both the repair volume that is missing from the damaged parts which should be restored by AM and the excess geometry which should be removed by subtractive machining. The proposed approach was implemented on three CAD models and three scanned models. The repair volumes for these damaged parts were successfully reconstructed. Three components were repaired via the directed energy deposition AM process. Finally, the bonding performance between the deposited layers and base parts was evaluated via tensile testing.


Mechanical and Aerospace Engineering

Keywords and Phrases

Additive manufacturing; Component repair; Damage extraction; Geometry reconstruction; Voxel-based modeling

International Standard Serial Number (ISSN)

2288-6206; 2198-0810

Document Type

Article - Journal

Document Version


File Type





© 2021 Springer, All rights reserved.

Publication Date

01 Nov 2021