| dc.description.abstract | Fused Filament Fabrication (FFF) is predominantly a rapid prototyping method, limited by its anisotropy and consequent unpredictability in component life. Due to unpredictability, researchers and industry do not consider FFF reliable for production of load-bearing components. In this Ph.D. thesis, I aim to identify the current state of research on FFF in the load-bearing application and contribute to understanding the process. In this thesis, I have identified four different levels of anisotropy in the FFF process, were layer bonding is considered the weakest axis with the highest failure rate. By focusing on the most critical element of the process, in this thesis, I propose that by using thermal data, I can analyze and predict the material properties of this weakest axis. The process determined by hardware, software, and calibration also significantly impacts material properties in all axes, where factors such as e.g., speed and acceleration cause anisotropic properties by changing fiber orientation and thermal profiles. In current research, hardware and manufacturing capabilities is a major limitations as current commercially available hardware does not provide desired manufacturing capabilities and sensors for research use. Therefore, in this Ph.D. thesis, I have contributed by creating an open-source hardware design that outperforms the datasheet in terms of UTS and is capable of producing all commercially available high-performance polymers by FFF. Finally, using our open-source hardware design, I have introduced Thermal Layer Design (TLD) and mapped out the correlation between UTS and layer temperature of PolyMide PA6-CF. Also, the temperature-to- UTS correlation is further analyzed. By adding additional sensors to an open-source printer, I first show how different geometries affect layer temperature in a component. Second, I map out the fracture surface using a temperature mesh of the printed samples. Using the knowledge presented in our previous contributions, I identify the potential of High-Performance Polymer Fused Filament Fabrication (HPPFFF) by doing complete material characterization by both process and mechanical properties, including fatigue of the polymer with the best initial potential, Polyether ketone with carbon fiber re-enforcements (PEEKCF) for use in component production.
However, creating components with the identified process parameters is still a limitation that needs further investigation in addition to utilizing the full potential of PEEK-CF, as shown by Injection Molding (IM). If fully optimized, HPPFFF, PEEK-CF has the potential to outperform most conventional structural materials, even titanium grade 6, in terms of relative strength, UTS divided by weight. PEEK-CF offers both rapid manufacturing and complex component production. There are still many unanswered research questions that need to be addressed to establish HPPFFF as a well-recognized production method but to get there, equipment manufacturers, companies, and researchers must understand the potential to invest in the idea. The main benefit of HPPFFF is the opportunity for added complexity, found in cases such as; Formula 1, manufacturing in and for space, and high-performance sports equipment. Therefore, in this thesis, I contribute to Paralympic sports by highlighting how to create big data for truly optimal equipment to enable athlete performance. The design of equipment optimized for individual human performance in a cyclic, power-generating motion is highly complex, and in addition to designing by motion capture, 3D scanning, and experiment design on geometries, iteration, and testing are still required. By hacking a commercially available printer, I produced large-scale components in hours, enabling triple-loop learning with a human-centered design. The result of this thesis, in addition to academic contributions, has increased the power output of the world record holder in Paralympic PR1 rowing by 47,6% in terms of power by equipment design and optimized power to propulsion, contributing to four world records. | en_US |
