Process-aware optimisation of lattice structure by electron beam powder bed fusion

Lattice structures are 3D open topologically ordered geometries that repeat an elementary cell in a predefined 3D space. Struts connected in specific nodes define the cell. Lattice structures are typical geometries that represent the design freedom unlocked by additive manufacturing (AM) and are unachievable with traditional processes. By tuning the morphometric parameters of the cell, its mechanical response can be significantly altered. Because of that, an accurate understanding of the process capabilities is crucial for achieving the nominally designed properties. Considering an electron beam powder bed fusion process, in this work, the same nominal lattice structure is produced under different processing conditions to determine the relationship between the process parameters, the actual cell morphometric parameters, and its mechanical response. Strut dimension, relative density and cross-section are measured using advanced X-ray computed tomography scanning analyses. Uniaxial compressive tests describe the mechanical performance. Inferential and descriptive statistical analyses are applied to investigate the effect of process parameters on the actual strut dimension and infer regression models. The results show that even slight variations of the process parameters significantly affect the morphometric structure parameters that result deviated from the nominal ones. The work demonstrates a strong correlation between all morphometric structure parameters and corresponding mechanical properties. The obtained regression model can predict the strut dimension from the process parameters, which can be then used to estimate the actual relative density and strut size. With this control and without any complex design procedure, a fine-tuning of process parameters allows a precise 3D spatial and localised control of structure properties to produce functionalised structures directly.