Learning from the Genoa bridge collapse

The tragic collapse of Genoa’s Morandi Bridge is a stark reminder of the risks of time-related deterioration and increased loading. Steve Jones, from the University of Liverpool, discusses the lessons learned.

[edit] Introduction

On 14 August 2018, one of the cable-stayed towers and adjacent spans of the Morandi Bridge in Genoa, Italy, collapsed, killing 43 people. The collapse occurred during adverse weather conditions, with reports that lightning hit the tower.

At the time of writing, we don't know the cause of the collapse and I'm reluctant to speculate until the facts are established and the official report is available.

The Morandi Bridge, constructed 1963–1967, was considered an advanced form of construction for its time. Morandi had previously designed the Maracaibo Bridge in Venezuela to a similar structural form.

The innovations of the Morandi Bridge were the early use of pre-stressed concrete and cable-supported spans. Over the past half-century, we have learned a lot about the use of these technologies. Two particular design aspects that bridge engineers of today are much more aware of are durability and robustness.

[edit] Durability and robustness

ICE journal papers on the collapse 33 years ago of the Ynys-y-Gwas Bridge in South Wales taught bridge engineers about the importance of making pre-stressed concrete bridges easier to inspect and to ensure that the steel pre-stressing components were adequately protected against corrosion.

Robustness is now a priority in any structure. Modern cable-stayed bridges usually adopt multi-cable arrays, which offer alternative load paths in the event of a loss of capacity of any one cable.

While these developments should make a tragedy such as the one in Genoa less likely, bridge engineers must continue to be aware of time-related deterioration and increased loading conditions.

ICE’s Bridge Engineering journal, of which I am deputy panel chair, continues to play its part in disseminating the latest international research and practice on bridge design, construction and maintenance. The latest issue is no exception, covering a wide range of topics from inspection and assessment of in-service bridges to innovative new designs.

[edit] Improved inspections

Desnerck et al. (2018) present a comprehensive study of the history, reported problems and current inspection regimes of reinforced-concrete bridges that include half-joints in England.

The half-joint was a popular feature of reinforced-concrete bridges in the 1960s and 1970s, but once the potential durability and maintenance problems associated with them were realised, their use was discontinued.

The paper describes the inspection and visual assessment techniques put in place by Highways England to manage the risks of its stock of 424 reinforced-concrete half-joint bridges that are still in use. Reference is made to how other nations' highway authorities are managing their stocks of similar bridges.

Griffin and Sivaji (2018) go on to describe the methods used to manage the assessment of the UK's huge stock of railway bridges owned by Network Rail.

The original three levels (1, 2 and 3) of bridge assessment are described along with measures taken to tackle the pressing need to quickly assess many thousands of bridges in a short time period. The development and introduction of a new rapid level 0 assessment is described.

[edit] Better modelling

Esteves et al. (2018) describe a rational design model for the connection regions between concrete and steel composite sections of bridge decks. The use of lighter steel composite decks in longer main spans connected to heavier concrete slabs in the side spans is becoming increasingly popular. The authors propose an improved method for the design of these difficult connections, illustrated with a helpful case study assessment.

The assessment of the dynamic behaviour of railway bridges often shows a marked difference between theoretical calculated behaviour and that measured by inspection. Typically, this would be in the form of lower measured natural frequencies than calculated by the structural analysis. This is often attributed to additional stiffness provided by the ballast and the track that are usually not included in the strength model.

Bigelow et al. (2018) propose a method for making allowance for the additional stiffnesses. A potentially useful formula for the quick assessment of the fundamental frequency of single-span railway bridges is proposed. Also on railway bridges, Augusthus-Nelson et al. (2018) describe experimental and numerical analysis studies of the influence of railway loading on soil-filled masonry arch bridges.

[edit] Innovative designs

Looking to the future, Zhang et al. (2018) describe the design and construction of the Ohmi–Odori Bridge in Japan. This is a 495m-long, four-span bridge with a multi-cell prestressed composite deck construction. Extradosed cables support the two longest spans (140m and 170m). It’s claimed to have the world's first extradosed bridge deck with corrugated steel webs.

The precast concrete FlexiArch system, previously described by Long et al. (2013), now has a wide range of applications for both new and replacement bridges. Long et al. (2018) now describe a further development of the use of the FlexiArch system for cost-effective widening of an existing masonry arch bridge in Bristol, UK.

This article was written by Steve Jones, senior lecturer in structural engineering, University of Liverpool. It was originally published by ICE on 3 January 2019 at https://www.ice.org.uk/news-and-insight/the-civil-engineer/january-2019/learning-from-the-genoa-bridge-collapse

--The Institution of Civil Engineers

[edit] Related articles on Designing Buildings Wiki

• Bridge construction.
• Bridges of the world.
• De la Concorde overpass collapse.
• Genoa bridge collapse.
• ICE articles on Designing Buildings Wiki.
• The Institution of Civil Engineers.
• Prestressed concrete.
• Reinforced concrete.

Comments

The assessment of the dynamic behaviour of railway bridges often shows a marked difference between theoretical calculated behaviour and that measured by inspection. Typically, this would be in the form of lower measured natural frequencies than calculated by the structural analysis. This is often attributed to additional stiffness provided by the ballast and the track that are usually not included in the strength model.

Increased stiffness would generally lead to higher natural frequencies, and high mass to lower natural frequencies. Perhaps there is a typo in the above paragraph?