Segmentation and Characterization of Surface Features to Aid Computational Modelling of Nickel Superalloy 625 As-Built Laser Powder Bed Fusion Parts

Published

July 11, 2022

Author(s)

Jason Fox, Christopher Evans

Abstract

The lack of strong correlations between the as-built surface topography of AM parts and heat transfer efficiency inhibits development for these applications. This issue stems from the complex and multi-scale surface topography along with significant variability in builds when compared to traditional fabrication methods. Correlation development can benefit from computational thermal and fluid modelling, but the large range of relevant scales on as-built surfaces make computation times impractical. To address this, data from a publicly available dataset of nickel superalloy 625 surfaces built in a laser powder bed fusion system are analyzed. Advanced characterization techniques and a novel segmentation technique are used to identify key features relevant to heat transfer. Analysis of the areal parameter Str shows correlation to the particle content of a surface. Particle segmentation creates attenuation of amplitude wavelength plots, particularly at the short wavelengths. Gaussian filters can be used to remove discontinuities and particles can be added arithmetically back to the surface based on segmentation data. Further simplifications of the surfaces are also presented. These methods provide simplifications that will reduce computational load and a foundation for experimental and modelling studies to come.

Citation

CIRP Annals

Pub Type

Journals

Download Paper

Local Download

Keywords

Additive manufacturing, laser powder bed fusion, surface metrology, segmentation

Citation

Fox, J. and Evans, C. (2022), Segmentation and Characterization of Surface Features to Aid Computational Modelling of Nickel Superalloy 625 As-Built Laser Powder Bed Fusion Parts, CIRP Annals, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=934095 (Accessed October 27, 2025)