With growing demand for air transportation, it is crucial to improve the efficiency of aircraft systems design. A key area of focus is the implementation of manufacturability considerations into the design process at early stages. This could be done by integrating manufacturability analysis modules into an MDAO workflow, however, each production method places unique requirements on the system. This study investigates the impact of considering manufacturing in the initial design phase, using a novel dynamic MDAO workflow that is capable of changing its analysis disciplines, design variables and constraints based on the current design point within an optimisation. This methodology was tested on a wing rib design case, with two production methods available: machining for metal ribs and stamp-forming for composite ribs. Integrating DFM analyses into the design study improved the trustworthiness of optimum results, removing non-manufacturable rib designs. Dynamic MDAO workflows yielded faster multi-objective optimizations and efficient handling of correlated objectives compared to conventional static workflows while providing similar results.

System architecting is one of the first stages of the engineering problem-solving process. Pivotal decisions regarding the system's overall configuration are taken in this phase. Consequently, decision support tools like system architecture optimization are needed to effectively assess the architectural design space. However, system architecture optimization problems include several features that make them challenging to tackle with traditional optimization techniques, including multiple objectives and mixed-discrete, hierarchical design spaces.

This work examines three algorithms capable of handling the mixed-discrete and multi-objective nature of system architecture optimization problems: genetic algorithms (such as NSGA-II), Bayesian optimization (BO), and the Deep Deterministic Policy Gradient (DDPG) reinforcement learning framework. These algorithms act as the engines that power three strategies capable of addressing hierarchical design spaces: \textit{global exploration}, \textit{nested optimization}, and \textit{decision chain}. While global exploration optimizes all design variables in a single loop and uses imputation on the inactive variables, the nested optimization method divides the problem into multiple hierarchical optimization loops. In contrast, the decision chain approach reframes the problem into a sequential decision-making process within an environment where an agent's actions alter design instances.

Test problems of varying complexity are used to evaluate the suitability of these algorithms and strategies. First, an airfoil optimization problem demonstrates the basic functionality of the selected algorithms with satisfactory results comparable to those found in the literature for the same problem. Next, a mixed-discrete version of the ZDT1 multi-objective problem is evaluated for different problem sizes and proportions of continuous and discrete variables. In this family of problems, the NSGA-II algorithm emerges as the most reliable and best-performing. At the same time, DDPG struggles with many variables, and Bayesian optimization runs into computational performance issues when dealing with large discrete combinatorial spaces.

Later, the architecture optimization strategies are evaluated on two versions of the Goldstein problem, a hierarchical and multi-objective test case. In the first version, which only considers eight architectures, all three strategies show competency in solving the optimization problem, but the global exploration strategy consistently shows the best performance. The decision chain strategy is discarded from further study at this stage due to relatively low performance and implementation difficulties. The second version of the Goldstein problem can be adjusted to different degrees of hierarchy and numbers of architectures. In this case, global exploration rises again as the best-performing strategy.

Finally, a real-world aileron structural optimization study at GKN Fokker shows that the implemented global exploration and nested strategies can find better designs than a random sampling of the design space. The best design for this test problem is achieved with a two-level nested strategy, combining NSGA-II in the outer loop and Bayesian optimization in the inner loop.

In conclusion, this research proposes three adaptive strategies powered by various algorithms to solve system architecture optimization problems. The findings suggest a preference for global exploration when a complete design vector can be pre-established. Otherwise, the nested optimization strategy, which can combine different algorithms' strengths and does not need a predefined design vector, emerges as the most suitable choice for more complex scenarios.

Requirement management is an important part of the design process. A perfectly feasible design can be useless if it does not meet the requirements imposed on the design. With the upcoming automation in the design process, requirement management can become even more complicated. How are requirements driving the design and how are these requirements verified? The requirement “trace” to the design can be unclear, which is the connection between the requirement and where the requirement is impacting the design. It is desired to know this as early in the design process as possible, since a design change due to a requirement compliance violation at the end of the design process is a costly matter. The aim of this research is to investigate how requirement traceability can be improved. This is done by developing a model-based requirement verification framework, which can automatically trace and verify requirements. This framework is also intended to formulate design workflows based on the requirements imposed on the design. The model-based approach is implemented and tested in a practical use case at Fokker Aerostructures in the front-loaded design of a wing box, as well as in a multidisciplinary design optimization problem of a simplified wing at Delft University of Technology. In both use cases, the use of the requirement verification framework led to a better understanding of how a requirement is impacting the system design and enabled the automatic generation of (optimization) workflows from a given set of requirements.

Considerations that decide producibility of a design are an important part of the design process, and must be included in the early design stages to ensure that these designs can be realised. Not including these considerations carries the risk of incurring additional costs and delays at later stages of product development because of design changes, or can lead to limiting oneself to conservative design choices to reduce the associated risk.
The current process in the industry accounts for these production considerations in design through a manual process that is iterative and time-consuming, and hence forms a bottleneck in being able to trade-off multiple design concepts. Attempts at accounting for these production considerations in an automated way are associated with the limitations of either only considering the manufacturing cost, being specific solutions that work only in certain scenarios, or being dependent on some commercial software tools, which are not fully suitable for use in context of automation and/or at the conceptual design stage. Additionally, the aspects of manufacturing and assembly are usually not considered at the same time in these studies.
Therefore, this thesis aims at developing a methodology that enables the automated inclusion of production considerations in the conceptual design process of aircraft structures, while overcoming shortcomings of the state-of-the-art....

The increasing complexity of aeronautical systems and the prevailing focus to develop sustainable technological innovations for future aircraft configurations and technologies have emphasized a particular demand on advancing the Multidisciplinary Design Analysis and Optimization (MDAO) technique. MDAO addresses the complex synergy within aircraft design where multiple coupled disciplines and domains need to be efficiently analyzed and optimized. Considerable developments within MDAO have already generated a significant time reduction for setup and solving of an MDAO problem and consequently in lead time and costs of aircraft design. For example, through the introduction of the software KADMOS (Knowledge-and graph-based Agile Design for Multidisciplinary Optimization System), that can automatically formulate MDAO workflows based on a repository of tools. The aims of this thesis are to further enhance the current application of MDAO for aircraft design and to alleviate the significant computational expense of MDAO problems through the consideration of sensitivity data during the MDAO formulation phase. Therefore, two approaches where the generation and evaluation of sensitivity information play a key role, were evaluated within this thesis, with the subsequent goal to shorten the MDAO execution time. Through the development and the following implementation of a sensitivity analysis tool into the sequencing and decomposition algorithms of KADMOS, this thesis demonstrated a new contribution to investigate on how to adjust the MDAO problem formulation. Sequencing is defined as the execution order and decomposition is the distribution of disciplines over various partitions. Furthermore, sensitivity information was used to identify the non-influential design variables respective to the objective during the MDAO problem formulation. Thus, during this thesis two independent methodologies were created in which the first methodology used a global sensitivity method to identify non-influential design variables, while the second methodology used a local sensitivity method to enhance the sequencing and decomposition algorithms of KADMOS. The methodologies were tested on an aircraft design problem and efficiently demonstrated the identification and removal of low sensitivity design variables from the problem formulation, with minor impact on accuracy. Overall, this thesis demonstrated that although incorporating sensitivity information in the MDAO problem formulation phase is advantageous, the computational effort to perform the sensitivity analysis may diminish the benefits.