Preprint
Article

This version is not peer-reviewed.

Fatigue Strength Estimation of PBF-LB/M AlSi10Mg Combining Confocal Measurements of As-Built Surface Defects with a √area-Based Approach

Submitted:

21 September 2026

Posted:

22 September 2026

You are already at the latest version

Abstract
Surface defects strongly influence the fatigue strength of AlSi10Mg manufactured by Laser Powder Bed Fusion for Metals (PBF-LB/M), particularly for specimens with as-built surfaces. This work presents a non-destructive procedure for determining a fatigue relevant defect parameter from confocal measurements of the as-built surface before fatigue testing. Surface valleys were identified from the measured three-dimensional topography and projected onto the plane perpendicular to the loading direction, which enabled the determination of the projected defect area A_⊥ and the corresponding defect parameter √(A_⊥ ), comparable to the parameter √area. The procedure was evaluated on 36 specimens covering variation in PBF-LB/M process parameters, different orientations of the specimens in relation to the building direction and two heat-treatment conditions, i.e., T6 and not heat-treated. After fully reversed axial fatigue testing (R=-1), the sizes of the crack-initiating defects were determined by fracture surface analysis and were compared to the values obtained by confocal microscopy. The differently determined defect sizes were generally in the same range, showing similar trends in dependency with the variation in manufacturing process and orientation of the specimens. Accordingly, the fatigue strength estimation by means of √area approach resulted in only small deviations (within a factor of 1.2) between the conventional approach based on fracture surface analysis and the methodology proposed in this work. These results demonstrate the applicability of confocal surface measurements for determining fatigue-relevant surface-defect parameters for non-destructive assessment of the fatigue strength.
Keywords: 
;  ;  ;  
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.