Additive manufacturing (AM) encompasses technologies that build part layer-by-layer, enabling the fabrication of near-net-shaped components with complex and customized geometries. In AM, the local heat source and processing conditions generate steep temperature gradients, leading to significant residual stresses. The distribution and evolution of these residual stresses depend on the process, part size, and geometry, as well as the microstructures and crystallographic texture of the material [1,2]. While typically undesired, textured microstructures are characteristic of AM and contribute to the enhanced mechanical properties, controlling them enables local tailoring of material behaviour.

Principal stress directions correspond to orientations where normal stresses are maximal or minimal and where shear stresses vanish, and are critical for understanding failure, yielding, and material response. Good understanding of principal stress magnitudes and directions as function of processing, geometry, and microstructure is therefore necessary to fully exploit the design freedom of AM. However, no consensus exists on expected principal directions or magnitudes, at the surface or in the bulk, due to the complex interplay of processing, material, microstructure, geometry, measurement location, and the limited number of studies to date [2]. Here we show that even in a simple ferritic AM cuboid with uniform texture, principal stress directions vary strongly with position, driven by boundary and free-surface effects.

To address this lack of understanding, we measured both the crystallographic texture and strain pole figures of a ferritic cuboid sample after AM at key locations: centre, face, and edge, via neutron diffraction at ENGIN-X (ISIS) beamline. Mantid software was used for texture analysis, while strains and stresses were determined using the inversion method proposed in [3], along with MTEX software.

Figure 1 shows the experimental (110) crystallographic pole figure and the corresponding (110) strain pole figures at two positions in the sample. The sample exhibits a strong cube texture aligned with the sample geometry [2]. The texture is uniform across the sample dimensions. In contrast, the strain pole figures differ markedly between the two positions: at the center position, the maximum strain lies at ~45° within the scanning plane—aligning with the diagonal of the sample—whereas at the corner position the maximum strain aligns with the building direction. These differences primarily reflect boundary conditions and free-surface effects, which significantly affect the magnitude and direction of the principal stress and ultimately part performance.

These results highlight the spatial variations in magnitude and stress distribution within an AM sample of simple geometry and uniform microstructure. Even in this straightforward case, the principal stress orientations are strongly location-dependent and may provide valuable prior information on their likely direction for any future time-constrained experiments. They also underline the critical importance of principal directions and texture in residual stress evaluation and the need to interpret measurements carefully when these factors are not fully considered.

[1] J., Schröder et al., Materials & Design, 2024, 244, 113171.
[2] S., Gaudez et al., Materials & Design, 2025, 251, 113658.
[3] M.A., Vicente Alvarez et al., Acta Materialia, 2024, 271, 119802.