学位プログラム名: 京都大学大学院思修館 / 京都大学 / 新制・課程博士 / 博士(総合学術) / 甲第23343号 / 総総博第16号 / 新制||総総||3(附属図書館) / 京都大学大学院総合生存学館総合生存学専攻 / (主査)教授 山敷 庸亮, 教授 寶 馨, 准教授 浅井 歩 / 学位規則第4条第1項該当 / Doctor of Philosophy / Kyoto University / DFAM

No description available.

No description available.

付記する学位プログラム名: 京都大学卓越大学院プログラム「先端光・電子デバイス創成学」 / 京都大学 / 新制・課程博士 / 博士(工学) / 甲第24622号 / 工博第5128号 / 新制||工||1980(附属図書館) / 京都大学大学院工学研究科電子工学専攻 / (主査)教授 木本 恒暢, 教授 白石 誠司, 准教授 小林 圭 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

This thesis presents a model for the procedures of manufacturing Fused Filament Fabrication (FFF) components by calculating required process parameters using empirical equations. Such an empirical model has been required within the FFF field of research for a considerable amount of time and will allow for an expansion in understanding of the fundamental mathematics of FFF. Data acquired through experimentation has allowed for a data set of geometric characteristics to be built up and used to validate the model presented. The research presented draws on previous literature in the fields of additive manufacturing, machine engineering, tool-path programming, polymer science and rheology. Combining these research fields has allowed for an understanding of the FFF process which has been presented in its simplest form allowing FFF users of all levels to incorporate the empirical model into their work whilst still allowing for the complexity of the process. Initial literature research showed that Polylactic Acid (PLA) is now in common use within the field of FFF and therefore was selected as the main working material for this project. The FFF technique, which combines extrusion and Computer Aided Manufacturing (CAM) techniques, has a relatively recent history with little understood about the fundamental mathematics governing the process. This project aims to rectify the apparent gap in understanding and create a basis upon which to build research for understanding complex FFF techniques and/or processes involving extruding polymer onto surfaces.

This thesis presents a model for the procedures of manufacturing Fused Fila ment Fabrication (FFF) components by calculating required process parameters using empirical equations. Such an empirical model has been required within the FFF field of research for a considerable amount of time and will allow for an ex pansion in understanding of the fundamental mathematics of FFF. Data acquired through experimentation has allowed for a data set of geometric characteristics to be built up and used to validate the model presented. The research presented draws on previous literature in the fields of additive manufacturing, machine engi neering, tool-path programming, polymer science and rheology. Combining these research fields has allowed for an understanding of the FFF process which has been presented in its simplest form allowing FFF users of all levels to incorporate the empirical model into their work whilst still allowing for the complexity of the process. Initial literature research showed that Polylactic Acid (PLA) is now in common use within the field of FFF and therefore was selected as the main working mate rial for this project. The FFF technique, which combines extrusion and Computer Aided Manufacturing (CAM) techniques, has a relatively recent history with lit tle understood about the fundamental mathematics governing the process. This project aims to rectify the apparent gap in understanding and create a basis upon which to build research for understanding complex FFF techniques and/or pro cesses involving extruding polymer onto surfaces.

No description available.

Material Extrusion additive manufacturing (MEX) represents the most widely implemented form of additive manufacturing due to its high performance-cost ratio and robustness. Being an extrusion process in its essence, this process enables the free form fabrication of a wide range of thermoplastic materials. However, in most typical MEX processes, only amorphous polymers are being used as feedstock material owing to their smaller dimensional shrinkage during cooling and well-stablished process/structure/property (P/S/P) relationship. Semi-crystalline polymers, with their crystalline nature, possess unique properties such as enhanced mechanical properties and improved chemical resistance. However, due to the inherent processing challenges in MEX of semi-crystalline polymers, the P/S/P relationships are much less established, thus limits the application of semi-crystalline polymers in MEX. The overall aim of this thesis is to advance the understanding of P/S/P relationship of semi-crystalline polymers in MEX. This is accomplished through both experimental and simulation-based research. With a typical commodity semi-crystalline polymer, Poly (ethylene terephthalate) (PET), selected as the benchmark material. First, we experimentally explored the MEX printing of both neat and glass fiber (GF) reinforced recycled PET (rPET). Excellent MEX printability were shown for both neat and composite materials, with GF reinforced parts showing a significant improved mechanical property. Notably, a gradient of crystallinity induced by a different toolpathing time was highlighted. In the second project, to further investigate the impact of MEX parameter on crystallinity and mechanical properties, a series of benchmark parts were printed with neat PET and analyzed. The effect of part design and MEX parameter on thermal history during printing was revealed though a comparative analysis of IR thermography. Subsequent Raman spectroscopy and mechanical test indicated that crystallinity developed during the MEX process can adversely affects the interlayer adhesion. In the third project, a 3D heat transfer model was developed to simulate and understand the thermal history of MEX feedstock material during printing, this model is then thoroughly validated against the experimental IR thermography data. While good prediction accuracy was shown for some scenarios, the research identified and discussed several unreported challenges that significantly affect the model's prediction performance in certain conditions. In the fourth project, we employed a non-isothermal crystallization model to directly predict the development of crystallinity based on given temperature profiles, whether monitored experimentally or predicted by the heat transfer model. The research documented notable discrepancies between the model's predictions and actual crystallinity measurements, and the potential source of the error was addressed. In summary, this thesis explored the MEX printing of semi-crystalline polymer and its fiber reinforced composite. The influence of MEX parameters and part designs on the printed part's thermal history, crystallinity and mechanical performance was then thoroughly investigated. A heat transfer model and a non-isothermal crystallization model were constructed and employed. With rigorous validation against experimental data, previously unreported challenges in MEX thermal and crystallization modeling was highlighted. Overall, this thesis deepens the understanding of current semi-crystalline polymer's P/S/P relationship in MEX, and offers insights for the optimization and future research in the field of both experiment and simulation of MEX. / Doctor of Philosophy / Material extrusion additive manufacturing (MEX), also known as fused filament fabrication (FFF), is a popular form of 3D printing known for its cost-effectiveness and versatility in creating objects from plastic materials. Traditionally, MEX utilizes amorphous polymers because they are less prone to shrinkage and thus easier to print. However, semi-crystalline polymers, offer enhanced strength and chemicals resistance, yet they pose significant challenges in printing due to a limited understanding of their process/structure/property (P/S/P) relationships in MEX. This research aims to improve our understanding of P/S/P relationships of semi-crystalline polymers in MEX. The study utilizes a typical semi-crystalline polymer, Poly (ethylene terephthalate) (PET), as the benchmark material. The study begins with the exploration of the MEX printing of recycled PET (rPET) and its glass fiber composite, finding that with appropriate MEX parameters, both feedstocks are highly printable, and the incorporation of glass fibers substantially increased the strength of the printed parts. Subsequently, a comprehensive investigation regarding the intricate relationship between crystallinity development, mechanical properties, and the MEX printing process is conducted. Our research revealed that the MEX process and the design of the part both considerably affect the crystallinity of the final part, thereby influencing its mechanical properties. In the third chapter, a 3D heat transfer model is constructed to better understand and predict the temperature evolution of materials during MEX printing. Most importantly, the modeling results are rigorously validated against experimental data, showing promising results. However, it also reveals challenges in precisely predicting the temperature of parts under certain conditions. The research then evaluates the applicability of Nakamura non-isothermal crystallization model for MEX printing scenarios. It is found that this model underestimates crystallinity in MEX, primarily because it does not account for shear-induced crystallization, a critical factor in the process. This finding underscores the necessity for more advanced models that can effectively capture the complex dynamics of MEX. In summary, this dissertation significantly enhances our understanding of the behavior of semi-crystalline polymers in MEX printing. It sheds light on the complex relationship between the printing process, the structure of the material, and the final properties of the printed object. This work not only advances our knowledge in 3D printing but also paves the way for more sophisticated modeling approaches, optimizing the MEX process and expanding its potential applications.

En principe, un laser femtoseconde peut être utilisé pour distinguer simultanément plusieurs molécules différentes dans l'air. L'idée principale de cette thèse est la détection à distance et la distinction de traces de polluants, tels que le méthane, en utilisant une technique de spectroscopic non linéaire de fluorescence induite par filamentation. Un système laser femtoseconde mobile est utilisé pour détecter à distance des traces de méthane dans l'atmosphère à l'extérieur et en plein jour (avec forte radiation solaire). Également, une méthode comprenant un système d'optique adaptative fonctionnant en boucle fermée est investiguée en tant que moyen d'augmenter significativement le signal de fluorescence pour des applications de télédétection basée sur la filamentation.

This thesis is concerned with fused filament fabrication (FFF) of cellulose nanocrystal (CNC) and thermoplastic polyurethane (TPU) nanocomposites, focusing on preliminary optimization of a processing window for 3D printing of mechanically responsive composites and the influence of temperature on mechanical adaptivity, thermal stability, and rheology. CNC thermoplastic nanocomposites are a water responsive, mechanically adaptive material that has been gaining interest in additive manufacturing for 4D-printing applications. Using a desktop FlashForge Pro 3D printer, we first established a viable processing window for a nanocomposite comprising 10 wt% CNCs in a thermoplastic urethane (TPU) matrix, formed into a filament through the combination of masterbatch solvent casting and single screw extrusion. Printing temperatures of 240, 250, and 260°C and printing speeds of 600, 1100, and 1600 mm/min instituted a consistent 3D-printing process that produced characterizable CNC/TPU nanocomposite samples. To distinguish the effects of these parameters on the mechanical properties of the printed CNC/TPU samples, a design of experiments (DOE) with two factors and three levels was implemented for each combination of printing temperature and speed. Dynamic mechanical analysis (DMA) highlighted 43 and 66% increases in dry-state storage moduli values as printing speed increases for 250 and 260°C, respectively. 64 and 23% increases in dry-state storage moduli were also observed for 600 and 1100 mm/min, respectively, as temperature decreased from 260 to 250°C. For samples printed at 240°C and 1600 mm/min, it was determined that that parameter set may have fallen out of the processing window due to inconsistent deposition and lower dry-state storage moduli than what the slower speeds exhibited. As a result, the samples printed at 240°C did not follow the same trends as 250 and 260°C. Further analysis helped determine that the thermal energy experienced at the higher end printing temperatures coupled with the slower speeds decreased the dry-state storage moduli by nearly 50% and lead to darker colored samples, suggesting CNC degradation. Isothermal thermogravimetric analyses (TGA) demonstrated that the CNC/TPU filament would degrade at relative residence times in the nozzle for all the chosen printing temperatures. However, degradation did not eliminate the samples' ability to mechanically adapt to a moisture-rich environment. DMA results verified that mechanical adaptivity was persistent for all temperature and speed combinations as samples were immersed in water. However, for the higher temperatures and slower speeds, there was about a 15% decrease in adaptability. Optimal parameters of 250°C and 1600 mm/min provided the highest dry-state storage modulus of 49.7 +/- 0.5 MPa and the highest degree of mechanical adaptivity of 51.9%. To establish the CNC/TPU nanocomposite's use in 4D printing applications, shape memory analysis was conducted on a sample printed at the optimal parameters. Multiple wetting, straining, and drying steps were conducted to highlight 76% and 42% values for shape fixity and shape recovery, respectively. Furthermore, a foldable box was printed to serve as an example of a self-deployable structure application. The box displayed shape fixity and recovery values of 67% and 26%, respectively, further illustrating significant promise and progress for CNC/TPU nanocomposites in 4D-printed, shape adaptable structures. Further analysis of the effect of degradation during FFF of the CNC/TPU nanocomposite was conducted using rotational rheometry, Fourier-Transform Infrared Spectroscopy (FTIR), and polymer swelling experiments. A temperature ramp from 180 to 270°C showed a significant increase in complex viscosity (h*) at the chosen printing temperatures (240, 250, and 260°C). Moreover, h* of neat TPU suddenly increases at 230°C, indicating a potential chemical crosslinking reaction taking place. 20-minute time sweeps further verified that h* increases along with steady increases in storage (G') and loss (G'') moduli. From these results, it was hypothesized that crosslinking is occurring between CNCs and TPU. Preliminary characterization with FTIR was used to probe the molecular structure of thermally crosslinked samples. At 1060 and 1703 cm-1, there are significant differences in intensities (molecular vibrations) as the temperature increases from 180 to 260°C related to primary alcohol formation and hydrogen bonded carbonyl groups, respectively. The hypothesis is the disassociation of TPU carbamate bonds into soft segments with primary alcohols and hard segments with isocyanate groups. The subsequent increasing peaks at 1060 and 1703 cm-1 may indicate crosslinking of CNCs with these disassociated TPU segments. To quantify potential crosslinking, polymer swelling experiments were implemented. After being submerged in dimethylformamide (DMF) for 24 hours, CNC/TPU samples thermally aged for 15 minutes at 240, 250, and 260°C retained their filament shape and did not dissolve. The 240 and 250°C aged samples had relatively similar crosslink densities close to 900 mole/cm3. However, from 250 to 260°C, there was about a 36% increase in crosslink density. These results suggest that crosslinking is occurring at these printing temperatures because both CNCs and TPU are thermally degrading into reactive components that will lead to covalent crosslinks degradation. Additional characterization is needed to further verify the chemical structure of these CNC/TPU nanocomposites which would provide significant insight for CNC/TPU processing and 3D printing into tunable printed parts with varying degrees of crosslinking. / Master of Science / This thesis is concerned with the development of a processing window for mechanically adaptive cellulose nanocrystal (CNC) and thermoplastic polyurethane (TPU) nanocomposites with fused filament fabrication (FFF) and, evaluating the influence of elevated temperatures on the mechanical, thermal, and rheological properties of said nanocomposite. CNC thermoplastic nanocomposites are a water responsive, mechanically adaptive material that has been gaining interest in additive manufacturing for 4D-printing. Using a desktop 3D-printer, an initial processing window for a 10 wt% CNC in TPU was established with printing temperatures of 240, 250, and 260°C and printing speeds of 600, 1100, and 1600 mm/min. A design of experiments (DOE) was implemented to determine the effects of these parameters on the mechanical properties and mechanical adaptability of printed CNC/TPU parts. Dynamic mechanical analysis (DMA) suggests that combinations of higher temperatures and lower speeds result in reduced storage moduli values for printed CNC/TPU parts. However, mechanical adaptation, or the ability to soften upon water exposure, persists for all the printed samples. Additionally, there was significant discolorations of the printed samples at the higher temperature and slower speed combinations, suggesting thermal degradation is occurring during the printing process. The decrease in storage moduli and discoloration is attributed to thermal energy input, as thermogravimetric analysis indicated thermal degradation was indeed occurring during the printing process regardless of printing temperature. Using the parameters (250°C and 1600 mm/min) that displayed the superior mechanical properties, as well as mechanical adaptivity, shape memory analysis was conducted. The optimal printed part was able to hold 76% of the shape it was strained to, while recovering 42% of the original unstrained shape once immersed in water, indicating potential for shape memory and 4D-printing applications. Furthermore, a foldable box was printed with the optimal parameters and it displayed similar shape memory behavior, illustrating promise for CNC/TPU self-deployable shape adaptable structures. To further study the effect of degradation on the CNC/TPU system, melt flow properties, molecular structure, and polymer swelling were investigated. At the printing temperatures (240, 250, and 260°C), the complex viscosity of the CNC/TPU filament experienced an exponential increase, indicating potential network formation between the CNCs and TPU. Fourier-Transform Infrared Spectroscopy (FTIR) highlighted changes in the molecular structure for the CNC/TPU filament as temperature increased from 240 to 260°C, which suggests that chemical structure changes are occurring because of degradation. The hypothesis is TPU is disassociated into free soft and hard segments that the CNCs can covalently crosslink with, which can potentially be explained by the increases in the FTIR intensities relating to TPU and CNC's chemical structure. To further quantify potential crosslinking between CNCs and TPU, polymer swelling experiments were implemented. The results from these experiments suggest that increasing printing temperatures from 240 to 260°C will lead to higher degrees of crosslinking. Further investigation could yield the validity of this crosslinking and additional optimization of FFF printing with CNC/TPU nanocomposites.