Predicting surface roughness of overhang structures by registering and fusing sensing signals and morphology data in LPBF

View at publisher Summarise with VIRAT
Authors: Longchao Cao, Tao Zhang, Xufeng Huang, Yahui Zhang, Wang Cai, Jie Li
Cited by: 0
References: 51

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Abstract Surface roughness is critical for the quality of overhang structures manufactured by laser powder bed fusion (LPBF). In-situ monitoring of surface roughness is promising to improve the surface quality. However, the up-skin and down-skin of overhang structures are often covered with unmelted powder or dross, making it difficult to establish a reliable connection between monitoring signals and the surface roughness. To address this issue, a new spatial information (SI) method was proposed to register sensing signals to the monitoring print parts. Firstly, an off-axial in-situ monitoring system with a microphone and photodiode is developed to capture process signatures during the LPBF, and the SI is acquired by inferring the coordinates of the laser scanning path from temporal features in the optical signal. Then, the signal distribution maps were reconstructed by mapping signals to SI. Finally, the Gated Transformer Network was developed for predicting surface roughness within each small area of the printed parts (352 μm × 250 μm). The registered signals exhibited strong effectiveness in predicting the surface roughness. Additionally, the relationship between the registered signals from the sub-surface region, defined as the deposited zone within 0–1.16 mm below the overhanging surface, and the resulting surface roughness was evaluated. The results indicate that the model achieved optimal prediction accuracy when the signal width was set to 0.58 mm. This work first provides a new approach for evaluating the surface quality of complex geometries.

Fields of Study & Subjects

Additive Manufacturing Materials and Processes Additive Manufacturing and 3D Printing Technologies Surface Roughness and Optical Measurements

Publication details

Journal / source
Measurement Science and Technology
Volume
37
Issue
12
Pages
125601–125601
Publication date
27 March 2026
Language
EN