Validation d’une méthode de contrôle par microtomographie à rayons X de la rugosité de surface de pièces obtenues par fusion laser sur lit de poudre à des fins de prédiction de performances

Abstract

Additive manufacturing, and more specifically the Laser Powder Bed Fusion (LPBF) process, has experienced sustained growth over the past several decades. This technology enables the production of components with highly complex geometries. However, this major advantage is accompanied by generally poor surface quality, making surface characterization essential before the deployment of manufactured components. Conventional surface inspection techniques are unable to access the complex geometrical features of LPBF parts, as they have primarily been designed for the inspection of easily accessible surfaces. To overcome these limitations, non contact, penetrating and non-destructive evaluation techniques, such as X-ray micro-computed tomography (μCT), have been developed. The objective of this master thesis is to validate one of these approaches in order to predict the in-service performance of LPBF components, particularly their fatigue life. To this end, the influence of acquisition resolution, surface finish, material density, and size effect on roughness metrics computed using an innovative surface determination method was quantified. To ensure a broad range of surface roughness values (Ra ranging from 10 to 30 μm), several benchmark artifacts featuring three planar surfaces oriented at 45°, 90°, and 135° relative to the build plate were fabricated in titanium alloy (Ti-6Al-4V) and nickel-based superalloy (IN625). An initial sensitivity study on acquisition resolution demonstrated convergence of the measured roughness values toward those obtained with a standardized instrument, regardless of build orientation or material density. A comparative study with conventional surface metrology techniques highlighted the excellent repeatability of the proposed method (Ra ± 0.4 μm) on a reference surface, placing it among the most repeatable measurement approaches. Furthermore, despite variations in material density and specimen size specifically intended to challenge μCT acquisitions, the proposed method consistently yielded results that were the closest to those obtained using the highest-resolution instrument, thereby demonstrating its accuracy. Overall, this master thesis establishes the foundations for a comprehensive investigation of the relationship between the surface texture of LPBF manufactured components, quantified through roughness metrics, and their mechanical performance. The rigorous validation of roughness extraction methods through comparison with established reference techniques constitutes the central theme of this work. Furthermore, it lays the groundwork for future studies involving more complex geometries representative of industrial applications.

Date	12 Aug 2026
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Vladimir Brailovski (Supervisor) & Jean-René Poulin-Masson (Co-supervisor)