Titolo della tesi: Advanced Design of Deployable Boom Structures
This doctoral research focuses on the development of advanced design methodologies and constituent material modeling tools for deployable composite booms (DCBs), with a particular emphasis on collapsible tubular masts (CTMs).
The design of DCBs presents a complex engineering challenge that requires a holistic approach to balance competing objectives while addressing demanding constraints. To tackle this complexity, this PhD project introduces innovative design frameworks capable of managing multiple variables, objectives, and constraints. These frameworks have been applied to typical DCB design scenarios.
One of the design scenarios addressed in this research focuses on the dual-state nature of DCBs, which must transition between compact and fully deployed configurations. To optimize both packaging efficiency and structural performance after CTM deployment, a design framework incorporating multi-objective optimization (MOO) and genetic algorithms (GAs) was developed. The MOO strategy enables the identification of compromise designs that single-objective optimization approaches might overlook, while the inclusion of GAs enhances the framework's adaptability. This versatility was demonstrated through the application of the framework to two distinct CTM architectures: one with a symmetric cross-section and one with an asymmetric cross-section. For each application, all benchmarked GAs performed well in terms of performance indices, such as the mimicked Inverted Generational Distance (mIGD) and the mimicked Hypervolume (mHV). Applied to the design scenario proposed in this thesis, the framework provides DCB developers with valuable guidelines, whether the objective is to prioritize bending properties along one principal axis or the other, or to emphasize compactness. If the goal is to prioritize bending stiffness around one axis over the other, an asymmetric cross-section is recommended, whereas a symmetric cross-section is more suitable when the priority is reversed. When compactness is the primary objective, an initial categorization of the optimal solutions obtained with this framework, based on the diameter occupied by the CTM in its coiled configuration, is suggested.
Another DCB design scenario considered involves studying the diverse load conditions that CTMs may encounter during their operational lifetime, with particular emphasis on buckling resistance under axial and bending loads. To enhance the buckling behavior of CTMs, the design framework integrates automated finite element analysis (FEA), neural networks (NNs), MOO, and GAs. Automated FEA facilitates the creation of a numerical dataset characterizing the buckling behavior of 1000 CTM samples under axial force and two orthogonal bending moments. The NNs develope surrogate models that extend the extensive but yet discrete FEA sample data into a continuous domain, providing excellent predictive performance with R^2 values of 0.9906, 0.9987, and 0.9628, for the axial force and the two bending moments, respectively. By solving a MOO problem using GAs and leveraging the quantitative information from NN-based surrogate models, the research uncovers correlations between improvements in buckling performance under the different loading conditions. These correlations provide designers with valuable insights, enabling them to select design spaces tailored to the anticipated operational loads of the CTM. Moreover, in addition to demonstrating efficiency in the analyzed case, the proposed design framework offers high applicability to various design scenarios, owing to the versatility and adaptability of the tools employed.
The dual-state nature of DCBs has been further explored to develop flexible CTMs that enable safe folding and stiff CTMs that ensure reliable operational performance, with particular focus on the interface between the boom and the hosting satellite. Commonly used interfaces include the clamped BC, wherein the CTM is cantilevered with its open root cross-section clamped to a planar surface, and the flattened and clamped boundary condition (BC), where the CTM is cantilevered with its root flattened and clamped. To propose a novel interface design, this study conducted extensive FEA to analyze the coiling and bending behavior of CTMs. Based on insights from the numerical results, a new interface design was developed, incorporating tape springs achieved through selective material removal at the CTM’s interface section. This innovative interface allows safe coiling around a hub, which would not have been feasible with the clamped BC, while achieving a bending stiffness 3 times greater than the flattened and clamped BC. By leveraging FEA, the design process significantly reduced development costs and time compared to traditional experimental trial-and-error methods. In addition to being cost-effective, the approach has proven highly effective. A lab-scale prototype of the new interface has been produced, and its folding behavior demonstrated consistency with simulation predictions during a safe coiling tests. These results underscore the utility of the proposed framework in developing optimized CTM interfaces and highlight the effectiveness of the novel design in enhancing both folding safety and operational reliability.
Another critical aspect highlighted in this research is material modeling. The focus has been on carbon fiber-reinforced polymers (CFRPs), which are commonly used for DCBs. Particularly, this research develops material modeling tools to better predict the characteristics of CFRPs that are crucial for DCBs development.
One of these characteristics is the out-of-plane behavior, which is critical due to the significant curvature changes that DCBs undergo. In particular, the research proposes a modeling tool to characterize the out-of-plane behavior of based-spread tow CFRPs, materials chosen for their advanced suitability for ultra-thin CTMs. The model uses homogenization techniques from the micro to macroscopic scale and approximates spread tow with very high aspect ratio (width-to-thickness ratio $\tilde 100$) as having a rectangular cross-section. This approach delivers highly accurate numerical results, validated against experimental data reported in the existing literature, with discrepancies of only 3% and 1% between simulation and experimental outcomes for 3- and 4-ply laminates, respectively. Moreover, it effectively captures the significant differences in out-of-plane behavior resulting from variations in thickness, all while maintaining a high level of simplicity. This model provides valuable insights that can inform material design during the DCB development phase.
Another key characteristic investigated in this research is the effect of radiation exposure, especially on the long-term behavior of CFRPs, considering the extended durations of space missions. Specifically, this study introduces a modeling tool to predict the impact of UV-exposure on the long-term behavior of unidirectional CFRPs. The model employs homogenization techniques from micro to macroscopic scales, approximating the time-dependent behavior of the constituent polymer using the Prony series. The numerical modeling results indicated that UV-C exposure led to only a slight reduction in the CFRP’s axial relaxation coefficient along the fiber direction, with no significant time-dependent degradation, as the fiber largely governs this behavior. In contrast, the axial relaxation coefficient perpendicular to the fiber direction, as well as the off-diagonal and shear relaxation coefficients, exhibited more substantial changes. Specifically, these coefficients showed an approximate 10% reduction in their initial values after UV-C exposure. Over a 32-year period, degradation became more pronounced, with differences between the pre- and post-exposure coefficient values reaching nearly 60%. These findings underscore the importance of considering both radiation exposure and mission duration during the DCB design phase to ensure the functionality of the structure remains intact. The proposed relaxation modeling tools are therefore considered valuable resources to be integrated into the DCB development process, ensuring the reliability and longevity of these structures.}
As of now, the design frameworks developed in this research have proven to be efficient in addressing the challenges associated with typical CTM design scenarios. In the near future, as technological advancements occur and the complexity of boom application scenarios increases, these frameworks are expected to remain highly relevant and increasingly valuable, thanks to their use of robust tools such as automated FEA, MOO, NNs, and GAs, which are applicable to a wide range of complex problems. Furthermore, the material models have proven to be provide essential insights about the critical characteristics of CFRPs for DCB applications, such as out-of-plane behavior of based-spread tows and relaxation behavior after exposure to harsh space conditions. By integrating these modeling tools into the development phase of DCBs, the efficiency and reliability of the proposed design frameworks is expected to be further enhanced.