People & Places

Strengthening of Historical Steel Bridges

21 June, 2022 12 min reading
Based on the dissertation by: Gary Ogden, M.Sc. in advanced masters in Structural Analysis of Monuments and Historical Constructions.

Strengthening of Historical Steel Bridges

The main aim of this thesis, by the author Gary Ogden, is to focus on the strengthening of historical steel bridges with an emphasis on the use of CFRP.




Bridges are important pieces of infrastructure for a variety of transportation methods and goods. From transporting passenger vehicles to freight trains, they are critical for social and economic activities by connecting two or more areas over an obstacle. Due to ever-changing loading and harsh environmental conditions, mainly due to the potential bodies of water they cross, bridges are in need of constant maintenance and rehabilitation. Unfortunately, due to their importance, the bridges are rarely completely closed for works. This leads to a significant amount of inadequate repairs with minimal impacts or even complete neglect.


Many steel bridges, especially in Europe and the Czech Republic, have been in use for an extended amount of time. Therefore, they are in need of significant repairs in order to continue with their intended function. The steel bridges focused on the thesis have the additional aspect to consider: they have been declared cultural industrial heritage.



NAKI II Project


The NAKI II project is a research program funded by the Czech Ministry of Culture in order to improve development and research of projects involving cultural heritage as well as national and cultural identities. The period of the program is eight years, starting from 2016 until 2022.


The basis of this thesis is from within the NAKI II program with a project entitled “The methods for achieving the sustainability of industrial heritage steel bridges”. It is focused on the development of methods and technologies for the diagnostics of steel structures based on corrosion and fatigue issues and the strengthening solutions for these on both joints and members.


The focus of the project is on CFRP, as it is a minimally invasive and reversible technique. Additionally, the project focuses on surface preparation, innovative painting systems with nanotechnologies, technical inspections and, finally, a case study application to a typical steel bridge. The latter objectives are not within the scope of this thesis.


Železniční most


The scope of this thesis will be limited to the examination of one joint detail of the Železniční most. The Železniční most, literally translated to “railway bridge”, is a steel railway bridge located in Prague, Czech Republic, just south of the old city. It connects the areas of Smíchov and Vyšehrad over the Vltava River crossing at Výtoň.


Constructed in 1901 with dual railway tracks and wooden pedestrian crossings at a cost of 4,000,000 Kč, it was built to replace the existing monorail bridge built in 1872. The bridge is a main link between Prague and the rest of the country, part of the “Prague connecting railway” for both passenger and freight services.



The new bridge is comprised of three semi-parabolic truss systems, each with a span of 72 meters, spanning 216 meters, supported by two stone pillars are located within the river. The trusses are made in the parabolic Bowstring style with a maximum height of 12.25m. Key features include the cut-off of the parabola at the ends and the three cross-counters at the center, compared to the single diagonals at the ends. The joints are made with riveted connections.



In 2008, the Czech Ministry of Culture declared the bridge an immovable cultural monument. This was decided as the bridge represents a highly valuable technical structure, exhibits late Classicism style and demonstrates the continuity of this building style into the beginning of the mid-20th century. The declaration as a cultural monument set back preliminary plans for building an adjacent bridge to satisfy increased traffic demands. 


Current State


The current condition of the bridge is that it is safe and useable for railway traffic. The wooden pedestrian walkways on both sides of the bridge have been closed since late-2017 to address safety concerns. 


Works to replace this part of the bridge are planned to take place is summer 2018 and 2019. Structurally, the bridge remains in fair condition after passing a load test, allowing the continued use of the bridge. The most concerning issue of the bridge is the significant corrosion present.



Addressing the corrosion of the bridge and its effect on the strength will be a main consideration presented in the thesis.





Traditional Methods


Traditional methods are ones that have been common throughout the history of steel bridge repair and typically feature alterations that align with the existing bridge technology of the original construction.


In general, traditional materials are well known, making them simple to work with and inexpensive. Typically, they don’t require specialized workers and have been rigorously proven in real structures and allow for existing codes to be applied. Drawbacks of these techniques include requiring a large amount of labor, having limited durably, and are very intrusive and bulky.



Steel Plates


Using steel plates as a strengthening method is very common due to both its availability and material properties.


A benefit of using steel plates for strengthening is that they are the same material as the existing members. Unfortunately, having the same properties also leads to drawbacks, such as corrosion. Steel plates add significant dead load to the structure compared to other options such as CFRP and, therefore, are more limited to localized strengthening applications.






Welding is a traditional technique that involves the fusion of two parts through the use of heat and/or pressure, commonly with an addition of a filler material. Once the heated material is cooled, the two parts effectively behave as one. Welding offers several benefits as a technique for applying the strengthening of steel plates. Welding is very dependent on the skill of the worker and welds must be inspected for quality and imperfections.





Bolting is one of the most common techniques in mechanical attachment of steel structures. It is a highly developed technique with a well-regulated manufacturing industry, resulting in numerous variation of bolting styles and confidence in their strength and available applications.





Riveting was the dominant mechanical joining technique for many structures from 1840s until 1940s and had a significant impact on the design of the structures. A rivet consists of two main parts: the rivet head, which is pre-manufactured, and the shank or shaft that is formed into the second head called the field head, when installed.





Replacement involves the removal of the original member(s) and installation of new members in place of the original(s). This can be limited to the changing of a single beam or members along with relevant connections. It is important to have an adequate design that does not change the load path significantly (through the new member’s increased strength), creating further damage to the original structure.



Additional Members


Adding members to an existing structure is another significant intervention and alternative to replacing a member by keeping the entire existing structure. The strengthening is a result of the new members sharing the load or taking over the load completely. In each scenario, the original load path is altered, which is a common critique of the method specifically for historical bridges, as the original structural system is not conserved.



Modern Methods


Modern techniques are the ones that have been recently developed for the use on steel bridge repairs. These innovative techniques are often created in order to overcome issues that the original construction techniques have. These include (but are not limited to): the increase of strength, reliability, reduced costs and less susceptibility to environmental factors. An important issue to consider with modern methods is the lack of knowledge of their long-term behavior compared to traditional methods.



Bonding Principals


In most structural applications, bonding involves attachment using an organic or synthetic polymer and relies on both adhesive and cohesive forces. Cohesion is the strength of the bond related to the interaction of the polymer within itself. Cohesive properties are designed and well controlled.




Steel Plate – Bonding


In addition to the traditional techniques of welding, bolting and riveting of attaching steel plate, bonding through chemical attachment is an option. There are many inherent benefits of bonding over aforementioned techniques.





Fiber reinforced polymer (FRP) is a composite material consisting of two constituents: a polymer that acts as a binder and for a fibre-matrix that provides the reinforcement. The polymer is also a protective layer and helps to transfer forces between the fibers. There are several commonly used fibers for FRPs including E-glass, aramid and carbon.






This technique involves the use of the aforementioned materials utilized in an active method. Pre-stressing is a concept where materials are subjected to tension forces an, in turn, induce compressive stress in order to directly counteract the tension forces present in typical use.



Steel Tendons


The use of pre-stressed steel tendons to apply post-tensioning to a structure is a versatile method of strengthening, where flexural, tension, shear, torsion, displacement and crack issues can be effectively addressed.



CFRP – Pre-stressing


Using pre-stressed CFRP strips builds on the benefit of CFRP having very high tensile strength. The use of pre-stressing is a more efficient method of strengthening, since it can act as an active reinforcement and transfer additional loads to the CFRP. This is especially true for steel structures, as steel is a relatively strong material and significantly more loading must be experienced by the structure before the CFRP is activated efficiently, compared to concrete structures.







The purpose of creating an FEA model is to supplement the state-of-the-art with numerical results of a real-world application by analyzing of a joint on the Železniční most.


These results can provide evidence for use in similar repair scenarios, as well as develop further understanding of the material interactions.


It also provides the visual and numerical comparison of methods and impact of the damage seen on the structure.


Additionally, the impact of various temperature and train loadings can be easily examined.


Finally, since wet layup CFRP systems are implemented only based on theoretical calculations, modeling can assist with a better understanding of the performance of the repair installations.




The FEA analysis of the selected joint does not include all the state-of-the-art methods described:


  • Replacement of members and additional members were excluded.
  • Pre-stressing of steel tendons was also avoided.
  • Bond interactions were also simplified in the analysis. 
  • The beam was analyzed assuming it behaved as a pure tension member. Changing forces as train pass or fatigue loading was not considered. Forces and boundary conditions were implemented in order to reflect only tension imposed on the member.
  • Welding is not a viable option in this strengthening scenario and, therefore, is not included.
  • For bolting and riveting, often the layout of the attachment will change. For example, using bolting typically will require less number of attachments compared to riveting. In order to limit further design of bolting, the existing rivet layout was used.



Selection of Joint Detail


The selected join detail to be modeled was a lower beam on the first span of the bridge. The location on the bridge is shown in Figure 3.1.



The selection was made as this particular joint had severe corrosion. The interior L-section webs of the joint have almost been completely interrupted. Therefore, it is a critical member to analyze and ideal for strengthening application. The overall member and close-up of the corrosion on this member is presented in Figure 3.2.



Model Generation


The model was generated using geometry created in AutoCAD and imported into Abaqus. It is a FEA software and the analysis was run using Abaqus/Standard, which uses implicit integration and an ideal for traditional modeling without severe nonlinearities.





The model represents one half of the lower beam. The lower beam is a built-up steel section consisting of Lsections on either side of two center plates. A base plate connects the two parts to create the entire lower beam. The base plate was modeled in its entirety.



Corrosion Modeling


In order to model the effect of the corrosion shown in Figure 3.2, a notch in the steel L-section was created. The outline of the loss of steel due to corrosion was made and then, subsequently, edited. The transformation of the “real” corrosion to “ideal” corrosion as well as the final section is shown in Figure 3.6.



CFRP Modeling


The CFRP was modeled using a shell element and assigned material properties with a composite section. The composite section comprises alternating layers of epoxy and CFRP sheet. The thicknesses are assigned in this section.



Steel Plate Modeling


Two attachment approaches were taken to model strengthening with steel plate mechanical attachment and bonding. Riveting and bolting models were combined. The steel plate was placed on the bottom of the L-section, as this is typical for installation purposes due to space restrictions for the alternative of installing on the top.





The mesh generated for the model used tetrahedral elements for solids and quadrilaterals for shells. In Abaqus these are identified as C3D10 and S4R, respectively. This notation describes that solid elements are created with ten nodes and shells by four.



Material Properties


Table 3.3 presents the material proprieties used for the model generation.





A state-of-the-art review on both traditional and modern techniques, for the strengthening of historical steel bridges, was presented.  Additionally, a FEA analysis was completed on a heavily corroded joint detail from the Železniční most in Prague.


  • The strengthening techniques involving member replacement or addition, as well as pre-stressing tendons, are not generally applicable. These techniques are generally considered to be too invasive and global approaches. Their use is limited to extreme damages and should be avoided if possible.
  • Bolting and riveting are viable options of attachment, with the latter more difficult to implement but more historically accurate. These attachment methods are severely limited to the geometry of the structure.
  • Welding of historical steels is difficult to preform, requiring extensive testing, at best, and is not possible at worst. Therefore, welding is also not generally applicable to use. 
  • Bonding eliminates the drawbacks from the other methods, but suffers from lack of extensive use and research.
  • CFRP, on the other hand, uses a similar bonding principal as steel plate bonding, but is significantly more developed. Additionally, its high strengthening-to-weight ratio, easy and adaptable installation process makes it more attractive for use over steel, making it the most promising option. As it is still a relatively new field, concerns with temperature effects and long-term behavior exist. Further research is still needed in order to increase the confidence in its use.
  • The FEA analysis provided further evidence for the use of CFRP as an ideal strengthening candidate. Three models were considered: steel bolting/riveting, steel bonding (with a thin and thick plate) and CFRP. 
  • The bolting/riveting model was not analyzed due to the need for additional plates and lack of space for attachment. This highlights, again, how limited by geometry this option can be.
  • Under normal loading conditions, results showed similar behavior between steel and CFRP, with the same strengthening ratio implemented. Differences arose when, due to geometry, the steel plate option created higher peak stress at the bond interface. This still occurred on the thicker steel plate, even with its higher strengthening ratio than the CFRP.
  • Subject to temperature change, the CFRP performed poorly compared to the steel plate. There was a clear increase in peak stresses with the CFRP model, whereas the steel plates exhibited a decrease.
  • Through the application of a minimal prestressing force, it is shown that the compressive forces can, effectively, be avoided improving the behavior of the CFRP strengthening method and addressing a main concern of CFRP systems. This is an important feature to acknowledge, as CFRP is able to be adapted to overcome drawbacks it possesses.
  • From the state-of-the-art and FEA analysis, CFRP is shown to be the suitable option for strengthening of historical steel bridges. This thesis provided evidence that the inherent benefits of CFRP heavily outweigh the drawbacks associated with the material. These drawbacks, particularly the influence of temperature, can be successfully addressed with proper design and use. Although more research is ongoing to understand the full behavior of the material, CFRP remains to be a promising solution compared to the other available techniques.



Future Work


There are several areas of focus available in order to further develop the results obtained from this thesis. The majority of them involve expanding the scope of the thesis in both the state-of-the-art and model.


It is possible to focus on fatigue loading and bonding for steel as well as CFRP in the state-of-the-art. 


Due to the size of the current model, it is well suited to be developed in two directions:


  • The global approach could involve the further accuracy of the model by detailing more of the joint with the addition of vertical members and the other half of the beam. This could be used to study the effects of unsymmetrical strengthening and force redistribution. With the focus on the bonding and fatigue loading, these could be applied to the model analysis as well.
  • Specifically, as the model currently assumes a perfect bond between the CFRP and steel, this could be modeled more accurately. With this improvement, the bonding could be studied in detail, addressing concerns with stress concentrations around the rivets, ends and corrosion notch. It could also be used to ensure the pre-stressing force does not cause any debonding issues or the need for anchoring.


Fatigue loading with the improved bonding accuracy would be another area of interest. In both models, the influence of the change of loading during a train-passing scenario could be developed. 


Finally, the exiting CFRP repair for which the model was based on should be monitored in order to observe long-term behavior.