(482f) Using Digital Twins to Model and Optimize Millifluidic, Multi-Material 3D Printing Nozzles

Multi-material additive manufacturing (MMAM) builds composite architectures to produce objects with enhanced functionality and mechanical properties. These structures can be reliably assembled using successive layer deposition, however, layer-by-layer MMAM suffers from low volumetric throughput as well as poor interfacial adhesion between layers. To address these challenges, the Bayles Group uses principles of advective assembly to engineer the next generation of MMAM printing nozzles. Advective assembly (AA) is a new processing technique [1] that structures composites by flowing distinct inks through a series of addition, rotation, and splitting elements. The elements sculpt laminar streamlines and template repetitive architectures such as stacks of layers and grids of fibers. By incorporating AA elements into a millifluidic nozzle, operators efficiently structure multi-material filaments before they arrive at the print bed. The modular combination of the flow elements allows operators to extrude voxelated, designer architectures provided that the flow remains stable.

Here, we use computational fluid dynamics to systematically investigate how ink rheology affects flow stability. Previous CFD studies have shown that when shear thinning, viscoelastic fluids are used in successive advective assembly operations, complex stress gradients arise and distort repetitive structures across the channel cross section.[2] In contrast, recent experimental studies [1] on granular microgel suspensions suggest that viscoplastic inks deform near the wall and flow as a stable plug elsewhere, improving structure fidelity. To systematically identify optimal operation conditions, we build digital twins of advective assembly nozzles in ANSYS Fluent 2023 R1. We simulate multiplicative flows of Newtonian and non-Newtonian fluids over a range of inlet velocities, yield stresses, power law indices, and viscosities. The effect of these parameters on the extruded structure is assessed by comparing the simulated output to a target repetitive architecture. Matrix comparison metrics (e.g. Jaccard index, matrix norm, and a modified Frobenius norm) quantify the magnitude of the distortion as a function of volumetric throughput. By identifying operational regimes that preserve structure fidelity, we aim to accelerate the optimization of millifuidic advective assemblers in MMAM.

[1] A. V. Bayles, et al., ACS Appl. Mater. Interfaces. 38, 21 (2022).

[2] P. D. Anderson, et al., Applied Rheology. 16, 198 (2006).