VOLCO: a predictive model for 3D printed microarchitecture

Extrusion-based 3D printing is widely used for porous scaffolds in which polymer filaments are extruded in the form of log-pile structures. These structures are typically designed with the assumption that filaments have a continuous cylindrical profile. However, as a filament is extruded, it interacts with previously printed filaments (e.g. on lower 3D printed layers) and its geometry varies from the cylindrical form. No models currently exist that can predict this critical variation, which impacts filament geometry, pore size and mechanical properties. Therefore, expensive time-consuming trial-and-error approaches to scaffold design are currently necessary. Multiphysics models for material extrusion are extremely computationally-demanding and not feasible for the size-scales involved in tissue engineering scaffolds. This paper presents a new computationally-efficient method, called the VOLume COnserving model for 3D printing (VOLCO). The VOLCO model simulates material extrusion during manufacturing and generates a voxelised 3D-geometry-model of the predicted microarchitecture. The extrusion-deposition process is simulated in 3D as a filament that elongates in the direction that the print-head travels. For each simulation step in the model, a set volume of new material is simulated at the end of the filament. When previously 3D printed filaments obstruct the deposition of this new material, it is deposited into the nearest neighbouring voxels according to a minimum distance criterion. This leads to filament spreading and widening, which is studied experimentally to validate the method. Experimental validation demonstrates the ability of the VOLCO to simulate the geometry of 3D printed filaments. In addition, finite element analysis (FEA) simulations utilising 3D-geometry-models generated by VOLCO demonstrate its value and applicability for predicting the mechanical properties of porous scaffolds. The presented method enables scaffold designs to be validated and optimised prior to manufacture. Potential future ...