Things bridges could know about themselves
Engineers develop models for calculating service life
by Julia Weiler
May 4, 2015
RUB engineers studied the flyover Pariser Straße at the junction Heerdter Dreieck in Düsseldorf for a period of two and a half years.
Traffic jam on the A 1 is currently an everyday occurrence. The bridge in Leverkusen, called “Rheinbrücke”, is clogged up; two lanes are closed, there is a speed limit of 60 kilometres per hour and a driving ban for heavy-duty vehicles with a legal maximum weight of more than 3.5 tons: safety measures, because the bridge is decrepit. It is in need of comprehensive repairs in order to remain serviceable long enough for a replacement to be built. Could the problem not have been spotted sooner? Perhaps in future, say the RUB engineers. They are developing mathematical models which can be used to predict the service life of building structures. The aim is an intelligent bridge. It would continuously monitor its own condition and raise the alarm once it detects signs of critical damage. Variations in temperature, wind, traffic, accidents and natural aging processes affect building structures in the long term. At the Institute of Concrete Structures, Dr Mark Alexander Ahrens and David Sanio analyse which factors are primarily responsible for fatigue of engineering materials and how precisely the service life of bridges can be predicted using mathematical models. For this purpose, they spent 2.5 years studying a flyover in Düsseldorf, namely “Pariser Straße” at the junction “Heerdter Dreieck” (fig. 1). Because the structure had been demolished and rebuilt, the engineers were able to have a good look inside and, as the lane was closed for traffic, to conduct tests immediately before demolition.
Using strain gauges, the engineers recorded over the period of eight weeks how steel deforms due to exposure to vehicle-caused strain.
One of their objectives was to refine pre-existing models for determining the service life of bridges. These models have been initially based on general assumptions that do not necessarily apply for each location in Germany. For example: the number of heavy-duty vehicles that cross a bridge per day varies from location to location. In calculations, such local differences have so far been replaced through general assumptions. The researchers from Bochum wanted to know how much more precise a prognosis could be if they tailor-cut the model for one specific structure. To this end, they recorded the current condition of the construction materials, congestion levels as well as environmental stress in Pariser Straße. For their tests, Ahrens and Sanio assessed the fatigue of high-strength pre-stressing steel only, not that of concrete. The examined flyover had been built as a pre-stressed bridge. The individual segments stick together, because they are compressed by steel tendons. It is as if one would hold a stack of books in front of one’s body by pressing the arms from left and right against it. Because of its lengths, the individual tendons of the structure are linked in several spots. “We know that segment joints are the weakest link in constructions of this type,” says Mark Alexander Ahrens.
Fig. 3© Wiley/Ernst & Sohn, in Sanio, Ahrens, Mark, Rode (2014): “Untersuchung einer 50 Jahre alten Spannbetonbrücke zur Genauigkeitssteigerung von Lebensdauerprognosen,” Beton- und Stahlbetonbau
Data recorded with a strain gauge over the period of four weeks; every day it recorded nine million measurement values. The graphic shows the total strain of the pre-stressed steel over several weeks (black graph) as well as the difference between minimal and maximal strain (grey bars). On weekends, especially on Sundays, Pariser Straße was less stressed than on weekdays.
In Pariser Straße, the engineers removed part of the concrete to expose the underlying steel – with the city’s permission, of course. They applied strain gauges to the exposed pre-stressing steel (fig. 2) and documented the extent to which the structure was affected by traffic over the period of several weeks (fig. 3). At the same time, they recorded a video clip in order to assign the observed effects to the respective vehicle types. Based on these data, Ahrens and Sanio identified the number of vehicles, including their loads, that drove over the bridge during the test period (fig. 4). They found out that Pariser Straße is used by fewer heavy-duty vehicles than expected. In this instance, the Germany-wide standard data had assumed excessive usage and, consequently, a shorter calculated service life than the flyover actually had.
Fig. 4© Wiley/Ernst & Sohn, in Sanio, Ahrens, Mark, Rode (2014): “Untersuchung einer 50 Jahre alten Spannbetonbrücke zur Genauigkeitssteigerung von Lebensdauerprognosen,” Beton- und Stahlbetonbau
Using a video clip, Sanio and Ahrens assigned load occurrences recorded by the strain gauge to certain vehicle types.
The high strain that a construction has to withstand determines its service life. But what about lighter vehicles? “Passenger cars are irrelevant for the calculation,” explains David Sanio. “Fatigue is caused by less frequent, yet higher strain.” In arithmetical terms, a heavy-duty vehicle affects the structure in the same way as 100,000 passenger cars. According to Ahrens, vehicles with a gross weight under five tons can be safely disregarded. What’s relevant are heavy-duty vehicles registered in Germany with a permitted gross weight of up to 40 tons, and first and foremost, heavy-haulage trucks. “Many routes are used daily by authorised heavy-haulage trucks that may carry a load of up to 250 tons,” says the RUB researcher. “They drive at night, when we are asleep, which is why we’re not aware of them. But for the bridge it doesn’t matter what time it is when it happens. The only thing that matters is that it does happen.” And it happens more and more frequently. The number of authorised heavy-haulage trucks has increased at an exponential rate in the last years; heavy-duty vehicles, too, drive more and more often. “It’s quite common for the vehicles to carry excessive load,” says Ahrens.
Fig. 5© Wiley/Ernst & Sohn, in Sanio, Ahrens, Mark, Rode (2014): “Untersuchung einer 50 Jahre alten Spannbetonbrücke zur Genauigkeitssteigerung von Lebensdauerprognosen,” Beton- und Stahlbetonbau
With a high-precision meter, a so-called leveller, the project team recorded by how many millimetres a bridge deforms when a heavy-duty vehicle stands on certain spots or drives over them.
In addition to traffic volume, temperature has also been a decisive factor for the service life of the flyover in Düsseldorf. Sanio and Ahrens measured it in different spots in the concrete over the period of several weeks and found out: sunlight hits the structure in irregular patches, which leads to a severe temperature gradient. The resulting tension inside the material overlaps with and magnifies the impact caused by heavy-duty vehicles. This result, too, shows how important it is to adapt the model to local conditions. The RUB team has refined it by adding more data. After Pariser Straße was decommissioned, the researchers, in collaboration with the municipality of Düsseldorf, examined to what degree the bridge is deformed by individual loads. To this end, they positioned a heavy-duty vehicle with a known weight in different spots and determined the extent to which the construction deflected with the aid of a precision leveller, i.e. an altimeter. This was an important quality-control measure for the calculation model. Ahrens and Sanio assessed how quickly material fatigue occurs in lab tests at RUB (fig. 5). They took steel samples which they pulled and then released several million times. The findings regarding material property gathered in this process were likewise fed into the model.
Fig. 6© RUBIN, photo: Gorzcany
This machine subjects the pre-stressing steel sample to fatigue stress: for how long can it be pulled and compressed before it breaks?
Using the refined model, the group from the Institute of Concrete Structures calculated once again how long the flyover in Düsseldorf would have still lasted. The results deviated significantly from the first prognosis that had been calculated using the general model. If they considered the local condition in their calculation, they found out that the actual service life would have been 14 times as long. Still, it’s too late to save Pariser Straße now. Due to comprehensive construction measures and traffic redirection at the junction Heerdter Dreieck, it does no longer exist. “The bridge has been demolished,” explains Mark Alexander Ahrens. “We took the trouble to drive out to the site on an early Sunday morning. But it was so foggy that we only heard a loud crash. We couldn’t see a thing.”
Thanks to the tests carried out by the Institute of Concrete Structures and in collaboration with other institutes, a fundus of algorithms has been compiled which describe the various factors affecting a structure. Engineers can use it like a modular kit to pick and choose the algorithms that are precisely suited for their individual application. In order to predict the service life of a specific bridge, a geometric model of the relevant structure has to be set up. The ultimate aim is to enable bridges to perform these calculations automatically one day. However, engineers have not yet accumulated any long-term experience regarding that measurement technology. For a structure to continuously monitor its own condition, temperature and strain meters as well as a meter to record individual traffic data would preferably have to be integrated during construction. “However, the current measurement technology has not yet been sufficiently tested in long-term applications,” explains Sanio. “It works well enough for weeks or months, but then uncertainty occurs.” Mark Alexander Ahrens adds: “An intelligent bridge has to be built in such a way that it not only maintains communication between the measurement unit and the engineer, but it also has to remain functional after 100 to 200 years. However, who knows, by that time we may well have found another solution for our transport operations and don’t need bridges anymore.”
The more you know: German bridges and Gigaliners – an impossible combination?
“Gigaliners weigh more than standard heavy-duty vehicles. But thanks to their superstructures, the individual axle loads are lower than those of the load models that we are applying today. It can be said in defence of Gigaliners that they would have a lesser impact than we assume in arithmetical terms. Consequently, I would not categorically say that Gigaliners should be rejected. It needs to be assessed on a case-to-case basis which routes are suitable for them and if any individual structures exist for which they would constitute additional load. It is a planning story.” Dr Mark Alexander Ahrens