GScan’s Muon Flux Technology deployed on an €18m bridge renovation project

Constructed in 1981, Sõpruse sild (Friendship bridge) is the longest publicly accessible bridge in Estonia, measuring 488.2m in total length. It is an extremely important transport corridor, connecting two parts of the city of Tartu.  

Renovation work is planned to take place until early 2027 to strengthen the bridge and extend its lifetime.   The contractors AS TREF and AS TREF Nord identified that the post-tensioning strands of the main girders of 33m spans were in poor condition. Therefore, the planned strengthening works might not be possible. Before they could proceed, they urgently needed a detailed assessment of the remaining steel inside the 12 bundles of prestressed strands.    

Muon Flux Technology was chosen to complete the assessment, as it is the only sensible investigation solution that can see through multiple layers of reinforcement.

With tested robust hardware, it was possible to deliver results quickly and to the required level of detail. Muon rays also penetrate deep into structures, making it ideal for assessing massive bridge structures.  

 Deployment  

This project required a rapid, on-site response. In collaboration with contractors AS TREF Nord, GScan's team swiftly installed a three-scanner setup, completing the process in just four hours. Our flexible, modular approach allowed us to quickly adapt to the site's unique requirements, demonstrating our efficient and targeted solution for complex infrastructure projects.  

The measurements lasted for 10 days, which was sufficient to collect the required amount of natural muon flux and to see the necessary bundles of steel strands. 

Measurement Results  

Since muons have a random direction and energy charge, in order to get good results, it is necessary to analyze the nature and difference in refractions of the flux and interpret the results based on statistical probability. 

Based on the raw data between two top and bottom scanners, it was possible to distinguish 10x different wire rope bunches in cross-section (Photo 4). In addition, 2x wire rope bunches were distinguished based on the analysis of the longitudinal cross-section (Photo 5) and metal elements, which are expected to be used to fix the position of the ropes (Photo 6). 

Photo 4. Cross-section of the bundles as seen from the direction of the S14 axis 

Photo 5. Longitudinal section when viewed from the outer side, where the two upper tensioning strand bundles and retainers can be seen separately 

Photo 6. Top view of the location of the attachment box and bundles of PT. 

In addition to processing the raw data, a machine learning metal detection algorithm was also applied, which was capable of automatically detecting rebars, stirrups and bundles of the wires (Photos 7 and 8). 

Photo 7. View of detected elements from the direction of S15. The missing steel and clearly distinguishable missing bundle of strands can be seen clearly at the bottom right edge close to the reinforcement. 

Photo 8. Side view of detected steel of the different bundles from the inside. Darker areas mark less steel and brighter areas more steel. It is possible to notice that one bundle (second from bottom) is hardly visible due to missing steel and at least five bundles are in relatively good condition. 

Thanks to the support of AS TREF Nord and the power of Muon Tomography, we were able to detect significant section loss in tendons, leading to a targeted, data-backed revamp to the repair scheme.  

Contact us for more information about deploying GScan’s MuFlux technology.