Computational Design Optimization of Large-Scale Building Structures: Methods, Benchmarking & Applications

Recent years have seen an explosion in the number of large-scale structures such as tall buildings and long span roofs (e.g. all but one of the world's 20 tallest buildings was constructed in the last 15 years, almost all of which are over 400m in height; furthermore, it has recently been estimated that 4 million skyscrapers of 40 stories will be required by 2050 to accommodate worldwide urban population growth). However, currently the forms of such structures are usually identified in an ad-hoc manner, with very limited application of optimization techniques, despite the fact that such techniques are now routinely used in other industrial sectors (e.g. automotive and aerospace). This means that material consumption and associated greenhouse gas emissions will often be far higher than necessary, and novel structural configurations which permit inclusion of energy efficient features such as light wells or atriums will often be overlooked.

In this project highly efficient mathematical optimization methods will be developed specifically for large-scale building structures, and used to automatically identify efficient layouts of structural elements. This will enable determination of the 'absolute minimum material reference design' for a given design brief, providing a powerful new means of evaluating the relative efficiency of alternative structural layouts. Methods will also be developed to automatically generate simpler and more practical structural layouts, which consume little more material than the absolute minimum quantity.

The methods will be used to identify structurally efficient layouts for a range of applications, including tall building exoskeleton design and long-span canopy roof design. Considering tall buildings, a recent development has been the use of exoskeleton 'diagrids', which give a clear expression of the structural system, and are perceived to be more efficient than conventional solutions. However, the use of any predefined configuration will implicitly inhibit efficiency and vast numbers of alternative layouts will be able to be considered using the tools to be developed in this project. Considering long-span canopy roofs, such as those used in sports stadia, exhibition halls and factories, reducing material consumption by adopting a more efficient layout of elements leads to a 'virtuous circle' since as structural self-weight is reduced, so does the amount of structural material required to support this.

The project will result in the development of practical tools and guidance for practitioners, and educational materials for students. Successful delivery of the research can be expected to dramatically improve the ability of engineers to design structurally efficient large-scale buildings.

This project will enable construction professionals to rapidly evaluate and optimize structural layouts for large-scale buildings, benefitting their clients and society at large, through lower cost, lower impact, buildings.

Structural engineering designers will benefit by having access to computational tools which allow the efficiency of proposed designs to be objectively judged, and new, highly efficient, designs to be automatically generated. In the short-term this will provide UK construction professionals with a competitive advantage, and will maintain their international reputation as pioneers of state-of-the-art technology. Their clients will also benefit by having access to more efficient designs, and a benchmark against which to compare design options, allowing them to commission buildings which are cheaper and/or have a lower carbon footprint, contributing to the UK government's target of halving the greenhouse gas emissions associated with construction by 2025. With compact connected cities regarded as a sustainable means of accommodating the rising global population, the number of large-scale buildings in the world is set to grow rapidly. Society at large will therefore also benefit from the research, through the availability of buildings which in future cost less and require less resources to construct.

Close cooperation with the collaborating practitioners will ensure that the methods developed are relevant and useful, and take account of both current and emerging fabrication techniques (e.g. additive manufacturing techniques, which allow bespoke 3D printed metallic joints, meaning that designs that have been difficult to fabricate in the past are unlikely to be so in the future). Practical guidance and demonstration software will also be developed for use by practitioners (see 'Pathways to Impact' document).

The project focuses on the design of large-scale buildings, where the form of the structure has a major impact on the volume of materials required. Also, because of the high loads in large buildings, bespoke elements are generally already necessary, so that there will often be relatively little overhead associated with adopting novel, highly optimized, designs. However, there is also scope for many of the techniques developed to feed into smaller-scale building design, using flagship large-scale building case studies to drive interest. Also, given the generic nature of many of the methods to be developed, these are likely to also benefit professionals involved in the design of components and structures in other sectors, such as bridge engineering, aerospace, automotive and space. Findings will be disseminated in sector-specific publications and/or software so as to maximise the potential for impact, leveraging existing links in these sectors (e.g. Gilbert is regularly invited to speak at UK bridge engineering conferences and is currently involved in the Innovate UK 'ANVIL' project, ref. 101483, involving aerospace, automotive and space sector industry partners).