Date of Award
Doctor of Philosophy
Brett G. Compton
Chad E. Duty, Dayakar Penumadu, Vlastimil Kunc
Over the past decade, the scale of polymer additive manufacturing has been revolutionized with machines that print massive thermoplastic parts with greater geometric complexity than can be achieved by traditional manufacturing methods. However, the heat required to print thermoplastics consumes energy and induces thermal gradients that can reduce manufacturing flexibility and final mechanical properties. With the ability to be extruded at room temperature and excellent compatibility with fibers and fillers, thermoset resins show promise to decrease the energy consumption, expand the manufacturing flexibility, and broaden the material palette offered by large-scale polymer additive manufacturing. However, structural instability in the uncured state has limited the scale of thermoset printing. This dissertation develops models to predict instabilities in printed thermoset structures and evaluates in-process chemically initiated curing to suppress instability.
First, instability is studied using direct ink writing to print thin walls with epoxy-based composite resins. Assuming these complex, viscoelastic resins exhibit elastic-plastic behavior, mechanical models based on measured rheological properties are shown to accurately predict collapse height, linking rheological properties to stability. This study helps estimate the scaling limits of thermoset additive manufacturing.
Second, unsupported, overhanging walls are printed with an epoxy-based composite resin. Again, assuming elastic-plastic behavior, an analytical beam bending model, finite element analysis, and rheological properties are used to predict the yield height and deflected profile of an unsupported overhanging wall. The predicted yield height and deflected profile agree with experimental observations. This study generalizes the link between rheological properties and structural stability, provides tools to analyze overhanging walls, and emphasizes scaling limitations.
Finally, a custom-built, large-scale printer is used to mix organic peroxide with vinyl ester resin immediately before deposition to initiate curing and suppress collapse. A stable processing window is revealed, where curing keeps pace with the rate at which layers are stacked on the print. Outside the stable processing window, successful printing is affected by pre-gel collapse, transient thermal behavior, and warpage. Preliminary modeling efforts are presented and motivate future research.
These studies advance large-scale thermoset additive manufacturing by providing ways to predict structural instabilities and evaluating one way to address instability.
Romberg, Stian K., "Structural stability of thermosets during material extrusion additive manufacturing. " PhD diss., University of Tennessee, 2021.
Video 2 - Thick FS-ink wall.avi (6240 kB)
Video 3 - Thin FS-ink wall.avi (2877 kB)
Video 4 - Thin NC-ink wall.avi (2417 kB)
8-16-21 Theta55 Try 4.mov (7149 kB)
Profile - alpha-5 alphaFE-4.8 Sample3.avi (4824 kB)
Profile - alpha-10 alphaFE-8.9 Sample4.avi (4364 kB)
Profile - alpha-15 alphaFE-13.9 Sample2.avi (3664 kB)
Profile - alpha-20 alphaFE-18.7 Sample1.avi (3471 kB)
Profile - alpha-25 alphaFE-23.3 Sample2.avi (3001 kB)
Profile - alpha-30 alphaFE-28.2 Sample3.avi (2793 kB)
Profile - alpha-35 alphaFE-32.6 Sample1.avi (2685 kB)
Profile - alpha-40 alphaFE-37.6 Sample3.avi (2252 kB)
Profile - alpha-45 alphaFE-41.6 Sample4.avi (2048 kB)
Profile - alpha-50 alphaFE-46 Sample4.avi (1822 kB)
Raw profile - alpha-0 Sample1.mov (9776 kB)
Raw profile - alpha-5 Sample3.mov (7779 kB)
Raw profile - alpha-10 Sample4.mov (7178 kB)
Raw profile - alpha-15 Sample2.mov (6936 kB)
Raw profile - alpha-20 Sample1.mov (6528 kB)
Raw profile - alpha-25 Sample2.mov (5993 kB)
Raw profile - alpha-30 Sample3.mov (5696 kB)
Raw profile - alpha-35 Sample1.mov (5190 kB)
Raw profile - alpha-40 Sample3.mov (4812 kB)
Raw profile - alpha-45 Sample4.mov (4483 kB)
Raw profile - alpha-50 Sample4.mov (4107 kB)
0_68 without fans - IR.mov (8792 kB)
0_68 with fans - IR.mov (5132 kB)
2_25 without fans - IR.mov (8339 kB)
2_25 with fans - IR.mov (13349 kB)
4_5 without fans - IR.mov (20366 kB)
4_5 with fans - IR.mov (21936 kB)
6_5 without fans - IR.mov (16076 kB)
6_5 with fans - IR.mov (10803 kB)