Key points are not available for this paper at this time.
In situ defect detection during electron beam powder-bed fusion (EB-PBF) remains challenging, especially for surface height measurement. Traditional surface height measurement techniques are typically conducted post-build and are limited to external surfaces. This study explores the use of backscattered electron (BSE) imaging for quantitative surface reconstruction in EB-PBF and examines its correlation to post-build porosity. A combined BSE ray tracing simulation–experimental approach was used to assess the influence of electron beam distribution and sampling resolution on reconstruction accuracy. Results show that BSE-based reconstructions are affected by point spread function of the electron beam, leading to underestimation of fine surface features. Simulated reconstructions generated by convolving optical microscopy data with a Gaussian beam kernel reproduced these effects, confirming the dominant role of the beam's point spread function. Applying Wiener deconvolution to the generated height maps significantly improved reconstructions. Layer-wise analysis demonstrated that variations in surface roughness metrics ( S a , S z , and S dr ) were correlated to process transitions such as column-to-overhang changes and reflect changes in melt pool stability. Comparison between deconvolved BSE reconstructions and post-build X-ray computed microtomography (μCT) revealed a moderate spatial correlation (Pearson Correlation Coefficient ≈ 0.45) between surface contours at −200 μm threshold and part porosity.
Upadhyay et al. (Mon,) studied this question.