Influence of print temperature and polymeric chain movement in interlayer development in fused filament fabrication

<p>The lack of current mechanical testing standardization in 3D printed parts means it is difficult to directly examine how the interlayer bond is developed through polymer thermodynamic movements. Presented in this thesis is an experimental study into the influence of molecular weight and print temperature on the interlayer strength in parts manufactured with Fused Filament Fabrication (FFF). Three PA12 grades with different molecular weight distributions (MWDs) were used to investigate the development of the polymer welded region between the deposited layers and had varying melt flow indexes (MFIs) denoted as (MFI 40.1, 4.08 and 0.55). These grades were initially printed into mechanical test samples at 200OC which revealed that the varying MFIs had significant impacts on the quality of the FFF mechanical test samples. Due to significant differences in viscosity between the PA12 grades, each material had distinct optimised print settings in terms of build orientation, layer height, print speed, and extrusion settings. Degree of Freedom (DoE) analysis of the tensile samples had shown that MFI alone is not significant, rather interaction between MFI, layer height and build orientation significantly impacted the maximum tensile strength. Due to process conditions, the impact of molecular weight on interlayer strength could not be determined through tensile testing directly. Instead, mode 1 fracture toughness with samples printed at 200OC, were able to reliably demonstrate an impact of MWD on the interlayer strength in terms of G1C values. The highest interlayer strength was observed in PA12 (MFI 40.1) which had the highest density of fast diffusing, low molecular weight polymer chains.</p> <p>These initial results developed into a more in-depth study into the interlayer bond formation with the inclusion of print temperature as a factor with the same fracture toughness samples mentioned above to be printed at 210, 230 and 250OC. The impact of this increased temperature decreased the calculated viscosity of the PA12 grades which resulted in lower porosity values and better interlayer contact in samples printed at higher temperatures and higher melt flows. Print temperature was effective in significantly increasing the fracture toughness values in each PA12 grade regardless of molecular weight and was supported by ANOVA statistical analysis. However, at increasing print temperatures, there was significant difficulty to track the crack propagation which did not demonstrate a clear G1C plateau and the ductility of the samples increased, causing extensive deformation of the arms and deviation from the standard used to conduct the fracture toughness test. PA12 (MFI 40.1) had particularly unstable crack propagation which meant it was difficult to obtain an average G1C value and had a high standard deviation. To prevent changes to the mode 1 fracture toughness sample geometry, the start G1C values (the energy used to break apart the initial crack) was used to indicate interlayer strength rather than traditionally using the plateau region (which did not exist). Despite difficulties encountered during measurement of this value, there was clear influence of MWD on the development of the interlayer region where PA12 (MFI 40.1) had the highest interlayer strength, followed by PA12 (MFI 4.08) and PA12 (MFI 0.55). ANOVA did not reveal a relationship between polydispersity (a measurement of MWD) and starting G1C values but this was because the calculation of the polydispersity term does not encapsulate the short lengths of polymer chains at the short end of the MWD spectrum which are likely to have diffused to create the interlayer weld region. Also, small-angle X-ray scattering (SAXS) and differential scanning calorimetry (DSC) investigations into crystallinity levels and structures revealed no discernible differences between the printed double cantilever beam (DCB) samples that could account for differences in interlayer strength.</p> <p>An investigation into how MWD and print temperature affects the welding time (time available for the weld region to form) and interlayer development was explored in relation to samples printed at 210, 230 and 250OC. Due to the semi-crystalline nature of PA12, bond development is different above melting point (Phase1) and below melting point (Phase 2) so bonding time and interlayer development was examined specifically within these phases during printing. In Phase 1, the PA12 grades exist in an amorphous state which has calculated diffusion rates and thermal profiles that indicated PA12 (MFI 40.1) had the most favorable interlayer development with fastest diffusion and slowest rate of cooling due to its high melt flow resulting in higher volume deposition. In Phase 2, bonding time was tracked with K type thermocouples with diffusion likely to be responsible for interlayer strength development compared to entanglement. A blended form of PA12 (PA12 (Blend)) was produced with equal ratios of PA12 (MFI 40.1) and PA12 (MFI 0.55) to understand the effect of a highly polydisperse structure on interlayer strength development in samples printed at 210, 230 and 250OC. Start G1C values increased with print temperature in a similar manner to the unblended PA12 grades but the interlayer strength of PA12 (Blend) surpassed PA12 (MFI 40.1) to have the highest G1C start value at the maximum bonding time which was attributed to the increased presence of low molecular weight chains and the potential presence of tie-chain molecules.</p> <p>The final chapter reported interlayer bonding in compression moulded (CM) samples through modified mode 1 fracture toughness tests. However, CM samples could not be successfully separated and were used as an indicator of maximum interlayer bonding. A modified DCB method was proposed where a new term “Degree of Healing (DH)” used for relative comparison of interlayer bonding to understand the level of welding in FFF manufactured samples. The DH calculation was based on initial FFF load divided by CM ultimate load where ultimate failure of the sample occurred. DH is relatively a new approach inspired by Wool/O’Conner equation which was mainly proposed for amorphous polymers. Several obstacles to consider this theory for semicrystalline polymers was discussed and the assumptions used to apply this theory. It was found that a general increase in DH corresponded with increased interlayer strength, validating that FFF G1C values were found to increase with increasing temperature and the amount of short polymeric chains within the MWD.</p>