Analytical, Experimental, Computed Tomography Inspection, and Finite Element Modeling of the Material Extrusion Wave Infill in Additive Manufacturing

Abstract

Material extrusion (MEX), also known as fused deposition modeling (FDM), is an additive manufacturing (AM) process that deposits a molten thermoplastic material layer-by-layer from a heated nozzle. This thesis investigates the use of MEX in the fabrication of a thin-walled structure (TWS), such as the ankle-foot orthosis (AFO). Three requirements for the AFO and other TWSs fabricated by MEX are that they are lightweight and durable and have tunable structural stiffness. To fabricate a lightweight TWS with a tunable structural stiffness, the layer-by-layer nature of the MEX process may be adapted to fabricate complex internal geometries within a part. The wave infill, which uses a sine wave pattern to fill the TWS, is one method for designing such an interior within a TWS. The key advantage of the wave infill is that its truss-like structure can minimize TWS mass and homogenize the TWS for characterization of its structural stiffness. The effect of the wave infill geometry on four metrics – stiffness, load capacity, fabrication time, and mass – was studied. Analytical models were developed that predicted these metrics to within 10% of experimental measurements. The analytical models were used to develop a composite simplification model (CSM) of the wave infill in TWSs with generalized geometries. In CSM, the wave infill and TWS faces are modelled as a homogenous stacked composite, which reduces computation and setup time. CSM for the wave infill was found to predict the stiffness of experimental measurements within 15%. An analysis performed on several geometries and loading conditions shows CSM to be a powerful finite element tool that can optimize the wave infill for TWSs.

To fabricate a durable TWS, interfacial weaknesses between layers due to voids from the MEX fabrication process must be inspected to allow for the improvement of MEX process parameters. Computed tomography (CT) is a non-destructive method for quantifying void density at MEX layer interfaces. An advanced segmentation technique called the Mixed Skew Gaussian Distribution (MSGD) method was developed to improve processing of CT AM part analysis. The MSGD method predicted the porosity of an AM specimen from the National Institute of Standards and Technology (NIST) to within 1% of its experimentally measured value. MSGD was applied to quantify the internal structure of a MEX filament and part. For the MEX part, average void area was found to be highest (>250 μm2) at the bottom of the layer and smallest (< 100 μm2) at the top of the layer, which could be explained by a large temperature gradient between layers and contractile thermal stresses inside the layer that causes the thermoplastic to have increased shrinkage resulting in larger voids. Overall, this thesis shows: (1) the wave infill can be used to generate a lightweight TWS with tunable structural stiffness, (2) CSM is a powerful finite element technique that may be used to design MEX wave infill TWSs, (3) CT and MSGD may be used to quantify the internal structure of MEX filaments and parts, and (4) voids from the MEX process occur at interfaces between layers, possibly due to large thermal gradients and plastic shrinkage. This research will inform and improve the MEX fabrication process to fabricate TWSs with tunable structural stiffnesses that are lightweight and durable.