A Synthetic Modal Generation of Additive Manufacturing Roughness Surfaces from Images

TL;DR

The paper addresses the challenge of generating representative roughness fields for CFD from limited additive-manufacturing surface data. It adapts Rogallo's Fourier-based synthetic turbulence method to create 2D surface-height fields from a single electron microscope image, yielding realizations characterized by a prescribed energy spectrum $E(|\mathbf{k}|)$ and scalable domain sizes $\tilde{L}_x$, $\tilde{L}_y$. Five Rogallo-based roughness realizations are generated and evaluated against the input via energy spectra and two-point correlation spectra, with the laser-path-parallel component $f_x^R$ providing the best anisotropic match. The approach avoids large image datasets typical of ML methods and produces grid-friendly roughness for CFD, though high-wavenumber differences arise from sampling and domain-size limitations. This method offers a practical, extendable alternative for AM roughness modeling in numerical simulations.

Abstract

A method to infer and synthetically extrapolate roughness fields from electron microscope scans of additively manufactured surfaces using an adaptation of Rogallo's synthetic turbulence method [R. S. Rogallo, NASA Technical Memorandum 81315, 1981] based on Fourier modes is presented. The resulting synthetic roughness fields are smooth and are compatible with grid generators in computational fluid dynamics or other numerical simulations. Unlike machine learning methods, which can require over twenty scans of surface roughness for training, the Fourier mode based method can extrapolate homogeneous synthetic roughness fields using a single physical roughness scan to any desired size and range. Five types of synthetic roughness fields are generated using an electron microscope roughness image from literature. A comparison of their spectral energy and two-point correlation spectra show that the synthetic fields closely approximate the roughness structures and spectral energy of the scan.

Figures (11)

• Figure 1: A schematic of the original and extracted image functions $[^{}]$ and $[^{}_{}]$.
• Figure 2: The structure of the synthetic field generation methodology and its results.
• Figure 3: The Fourier-Series approximation $[^{}]$ of the input roughness surface by Altland et al.AltlandEtAl2022 created with , = 64 period samples, plotted with a resolution of $[_{}] = 1950$, $[_{}] = 1186$ and scaled with the scaling factors $= 2\pi/14$ and $= 2\pi/8.5$.
• Figure 4: The -vector component synthetic roughness field $[^{}_{}]$, plotted with a resolution of $[_{}]$, $[_{}]$ = 500 and the domain scaling factors $= 2\pi/14$ and $= 2\pi/8.5$.
• Figure 5: The -vector component synthetic roughness field $[^{}_{}]$, plotted with a resolution of $[_{}]$, $[_{}]$ = 500 and the scaling factors $= 2\pi/14$ and $= 2\pi/8.5$.