Addressing the complexity of sustainability-driven structural design: Computational design, optimization, and decision making

Doctoral thesis, 2021

Being one of the sectors with the largest environmental burden and high socio-economic impacts sets high requirements on the construction industry. At the same time, this provides the sector with great opportunities to contribute to the globally pursued sustainability transition. To cope with the increasing need for infrastructure and, at the same time, limit their sustainability impacts, changes and innovation in the construction sector are required. The greatest possibility to limit the sustainability impact of construction works is at the early design phase of construction projects, as many of the choices influencing sustainability are made at that point. Traditionally, an early choice of a preferred design is often made based on limited knowledge and past experience, considering only a handful of options. This preferred design is then taken on to the successive stages in the stepwise design process, leading to suboptimization.

Alternatively, many different design choices could be considered and evaluated in a more holistic approach in order to find the most sustainable design for a particular application. However, finding design solutions that offer the best sustainability performance and fulfil all structural, performance and buildability requirements, require methods that allow considering different design options, analysing them, and assessing their sustainability. The aim of this thesis is to explore and develop methods enabling structural engineers to take sustainability objectives into account in the design of structures.

Throughout this thesis, a number of methods have been explored to take sustainability aspects into account in the structural design process. As a first step, highly parameterized computer codes for sustainability-driven design have been developed. These codes interoperate with FE analysis software to automatically model and analyse design concepts over the whole design space and verify compliance with structural design standards. The codes were complemented with a harmonized method for life cycle sustainability performance assessment, in line with the state-of-the-art standards. Here, sustainability criteria were defined covering environmental, social, economic, buildability and structural performance for multi-criteria assessment of design concepts. To identify the most sustainable designs within the set, multi-objective optimization algorithms were used. Algorithms that address the high expense of constraint function evaluations of structural design problems were developed and integrated in the parameterized computer codes for sustainability-driven design. To ensure the applicability and validity of these methods, case studies based on real-world projects and common structural engineering problems were used in this thesis. Case studies for bridges and wind turbine foundations as well as a benchmark case of a reinforced concrete beam were investigated.

The case studies highlight the potential of the methods explored to support the design of more sustainable structures, as well as the applicability of the methods in structural engineering practice. It is concluded that it is possible and beneficial to combine computational design, life cycle sustainability assessment, and multi-objective design optimization as a basis for decision making in the design phase of civil engineering projects. A wide adoption of such a sustainability-driven design optimization approach in structural engineering practice can directly improve the sustainability of the construction sector.