Generation of granular media through 3D printing: a comparative evaluation from the morphological perspectives

The generation of artificial granular media to investigate micro-to-macro correlations in sands is one of the innovations inspired by the recent advancements in 3D printing technology. While several 3D printing techniques exist to print granular particles, the basis for the selection of a specific technique and the relative accuracy in mimicking the morphological features are yet to be investigated. This paper investigates the accuracy of the reproduction of granular morphology by three widely used 3D printing techniques. Polyjet, Digital Light Processing (DLP), and Stereolithography (SLA) printing techniques are used to generate the analogues of reference sand particles of size range 1.76–6.39 mm. Subsequently, the 3D morphological indices of the printed grains are computed using X-ray micro-computed tomography (µCT) imaging followed by spherical harmonic (SH) particle reconstruction and computational analysis. These indices are compared with those of the reference particles, and the errors in the computed morphological parameters are obtained for the three different 3D printing techniques. The errors are found to be the lowest for polyjet-printed particles and the highest for SLA-printed particles. The accuracy of the reproduction of morphology is found to increase with an increase in the particle size.