Control of Process Settings for Large-Scale Additive Manufacturing With Sustainable Natural Composites
We present an additive manufacturing system for 3D printing large-scale objects using natural bio-composite materials. The process, affine to the Direct Ink Writing method, achieves build rate of 2.5cm3/s using a precision dispensing unit mounted on an industrial six-axis robot. During deposition the composite is wet and exhibits thixotropy. As it loses moisture it hardens and shrinks anisotropically. This paper highlights work on controlling the process settings to print filaments of desired dimensions while constraining the operating point to a region where tensile strength is maximum while shrinkage is minimum. Response surface models relating the controllable process settings such as Robot Linear Velocity, Material Feed Rate and Nozzle Offset, to the geometric and physical properties of an extruded filament, are obtained through Face-centered Central Composite Designed experiments. Unlike traditional applications of this technique which involve identifying a fixed optimal operating point, we use these models to first uncover the possible dimensions of a filament that can be obtained within operating boundaries of our system. Process setting predictions are then made through multi-objective optimization of the mathematical models. An interesting outcome of our study is the ability to produce filaments of different shrinkage and tensile strength properties, by solely changing process settings. As a follow up, we identify the optimal lateral overlap and inter-layer spacing parameters to define toolpaths to print 3D structures. If unoptimized, the material’s anisotropic shrinkage and non-linear compression characteristics cause severe delamination, cross-sectional tapering and warpage. Lastly, we show the linear scalability of our shrinkage model in 3D space which allows us to suitably compensate toolpaths to significantly improve dimensional accuracy of 3D printed artifacts.