Binder jetting of minimally processed, locally sourced sediments offers a promising route toward sustainable additive manufacturing, but powder spreading of poorly sorted, irregular grains remains underexplored. This study uses high-fidelity Discrete Element Method (DEM) simulations to quantify how process parameters govern layer quality when spreading an unrefined sandy silt from the Gypsum Hills, Kansas, USA, with a counter-rotating roller. Twenty-five multi-sphere particle shapes were reconstructed from morphological measurements and used to simulate up to 300,000 interacting grains while systematically varying roller traverse speed, rotational speed, diameter, and layer thickness. Traverse speed and layer thickness emerged as the dominant controls on powder bed density and uniformity: the slowest traverse speed (20 mm/s) produced the highest density (~ 0.50) and lowest variability (~ 0.08), and layer thicknesses ≥ 120 μm were required to avoid jamming and obtain consistent packing. In contrast, roller diameter and rotational speed had only minor influence over the explored ranges. The simulations also revealed percolation-driven segregation, with fine particles preferentially settling beneath coarser grains. Comparison with an idealized spherical powder showed that the silt is intrinsically harder to spread, achieving lower density (0.51 vs. 0.59) and higher variability. These results demonstrate that binder jet processing of unrefined silt is feasible but favors slower spreading and thicker layers that may reduce throughput and resolution, and they provide process design guidelines and particle-scale insight for deploying raw earth feedstocks in sustainable binder jetting.
Qabani et al. (Mon,) studied this question.