Innovative Design of Steel Girders in Cable-Supported Bridges: By application of numerical optimization methods

Research output: Book/ReportPh.D. thesisResearch

18 Downloads (Pure)


The main design principles for girders in cable-supported bridges have remained largely unchanged for the past 50 years and have reached a point where the potential for further development is limited. The design concept of closed steel box-girders with orthotropic stiffened decks has been used extensively in major cable-supported bridges due to its many advantages compared to the alternative of classic truss girders. However, the design concept is subject to substantial inherent fatigue issues. Furthermore, in future super-long bridges with spans beyond 3 km, the girder self-weight becomes a critical design factor preventing even longer bridge spans. Moreover, considering that the construction industry accounts for 39% of the world's CO2 emissions, attention must be broadened from the one-sided focus on construction costs to reducing material consumption.
To accommodate the challenges of decreasing self-weight significantly and reducing fatigue issues, it is anticipated that radical design changes will be required. With weak out-looks to new light-weight-high-strength materials, the identication
of new, innovative, and more material-efficient design concepts using existing materials is needed.
In this thesis, entitled "Innovative design of steel girders in cable-supported bridges", three different methods of structural optimization are applied in search of innovative girder concepts. The main focus is on reducing self-weight and on identifying more efficient load-carrying principles and, thus, more material-efficient structures.
Initially, parametric optimization is applied to pursue possible weight savings to conventional girder design. As a basis, a multiscale finite element model with sophisticated boundary conditions is established. Subsequently, a simple parameter study is carried out, followed by a gradient-based optimization with constraints on fatigue and deformation. The main findings are possible weight savings in the range of 6%-14%, achieved by using thinner plates and narrower stiffening troughs. Despite the possible weight savings, the results are considered modest, and it is confirmed that the conventional design concept is limited in further development,
without altering the structural concept.
Next, topology optimization of continuum structures is applied as the first step in search of innovative girder designs. The method is applied in giga-scale (2.1 billion finite elements) with a minimum of restrictions to identify a lower bound of the optimized designs. The highly detailed and intricate structure evolving is signicantly different from the conventional design, indicating more efficient loadcarrying 
principles. Based on the main structural features of the optimized designs, a simple interpreted design is established from where an initial weight saving of 13% is achieved. After a subsequent simple parametric optimization to identify the full potential, a total weight saving in excess of 28% is gained while maintaining manufacturability.
Finally, large-scale truss optimization based on finite element limit analysis is applied with constraints on stresses as well as global and local stability. A significant weight saving of 45% is achieved with a truss girder considerably different compared to the conventional design. Notably, a torsion grid evolves along the circumference of the domain, as well as large members in the bottom to carry loads primarily in tension. However, the higher weight savings gained with the truss girder, compared to the interpreted design, are achieved at the expense of increasing structural complexity, and thus, aggravated constructibility. Finally, potential weight savings of up to 54% are observed trough a simple parameter study.
The possible weight savings, identied from the various optimization studies, translate into total savings in material quantities of the entire bridge in the range of 16%-30%, and a reduction in CO2 emissions in the range of 18%-30%.
The identified design principles and possible weight savings emphasize the potential of using signicantly different girder design concepts. Hence, the potential weight reductions to be achieved may close the gap toward super-long cable-supported bridges and reduce material quantities, and thus reduce the environmental impact.
Original languageEnglish
PublisherTechnical University of Denmark, Department of Civil Engineering
Number of pages302
ISBN (Electronic)87-7877-524-8
Publication statusPublished - 2019

Bibliographical note

Report R-425

Fingerprint Dive into the research topics of 'Innovative Design of Steel Girders in Cable-Supported Bridges: By application of numerical optimization methods'. Together they form a unique fingerprint.

Cite this