Century of Data Reveals Weak Spot in Italy’s Aging Bridges and Tunnels

Italy’s motorway network is one of the great engineering achievements of the twentieth century, a lattice of reinforced concrete bridges and tunnels stitched across mountains and valleys during the country’s post-war economic miracle. But the very speed of that construction boom may have left a hidden weakness in the concrete itself. A new study drawing on more than half a million historical compression tests, together with over 4,500 cores drilled from in-service motorway bridges and tunnels, has produced the first decade-resolved picture of how concrete strength in Italian infrastructure has evolved over time, and it has uncovered a surprising anomaly: concrete cast in the first half of the 1970s is measurably weaker than expected, falling below the reference curves that engineers currently rely on to assess aging structures.

The research, led by Alessandro P. Fantilli and Mustapha Hatoum of the Politecnico di Torino and published in Results in Engineering, addresses a problem that has grown urgent since the catastrophic collapse of the Polcevera viaduct in Genoa in 2018. Most of Italy’s reinforced concrete motorway structures were built in the 1960s and 1970s, meaning many have already surpassed their design life of fifty years while remaining in daily service. Deciding whether such structures can be safely maintained, or must be expensively rebuilt, hinges on knowing the actual strength of the concrete inside them, a quantity that standards documents estimate only in the broadest of strokes.

Current international guidance offers little precision. The American codes ACI 562-25 and ASCE 41-17 provide tables of plausible compressive strength values spanning entire eras: for structures built from 1970 onward, for example, the suggested strength of beams ranges from 21 to 35 megapascals, while columns may span anywhere from 21 to 70 megapascals. These ranges are so wide, and their temporal resolution so coarse, that they can only support a rough assessment. For a national inventory dominated by structures from two specific decades, engineers needed something finer: a way to estimate, from the year of construction alone, what strength the concrete in a given bridge girder or tunnel lining is likely to possess, and whether it warrants closer investigation.

The Italian team’s answer came from an extraordinary archive. The Politecnico di Torino has been performing uniaxial compression tests on concrete since 1915, accumulating a database of more than 500,000 results, much of it stored as hard-copy records. Because digitizing every year of the century was impractical, the researchers selected one representative mid-decade year for each decade from 1915 to 2005, some containing tens of thousands of tests; the year 1975 alone contributed 21,672 results. Rather than assuming any theoretical distribution, they computed the 5th, 25th, 50th, 75th and 95th percentiles of strength empirically for each year, ordering the raw data and reading off the values at the corresponding ranks. The result is a set of strength-for-age curves, conceptually akin to the growth charts pediatricians use to track children’s development, that trace how typical concrete strength in Italy has risen across a century of changing materials and practice.

To validate these curves against reality, the team turned to the motorways themselves. Working with data shared by Autostrade per l’Italia, they analyzed 2,519 cores drilled from 135 tunnel linings built between 1961 and 1995 in north-west Italy, sampled at the two zones where damage typically concentrates: the crown and the spring lines. They also examined 1,981 cores from 58 reinforced concrete bridges built between 1958 and 1995 on two Italian motorways, grouped by structural element: main girders, transverse beams, deck slabs, pier elevations, pier caps and abutments. Cores damaged during drilling or showing defects such as honeycombing, embedded reinforcement or cracking were excluded on physical grounds, and size-correction factors from ACI PRC-214.4-21 normalized all results to a standard height-to-diameter ratio of one.

The comparison delivered a striking finding. Overall, the measured strengths tracked the rising trend of the century-long curves, and in many cases exceeded the conservative baselines of ASCE 41-17. Main girders, typically prestressed in the bridges studied, showed the highest values, averaging around 50 megapascals in the early 1960s and climbing toward 60 megapascals in later decades, with trendlines hugging the 95th percentile of the reference curves. But when the data were grouped by decade, the 1970s stood out for the wrong reasons. Tunnel concrete averaged 34.9 megapascals in the 1960s, dropped to 25.5 megapascals in the 1970s, roughly 30 percent lower, then nearly doubled to 55.1 megapascals in the 1990s. The same dip appeared in bridge transverse beams, pier caps and abutments, where early-1970s averages fell below even the 50th percentile of the strength-for-age curves.

The authors propose a plausible explanation rooted in the history of cement chemistry and construction practice. During the economic boom, the priority was speed: expanding the motorway network quickly demanded faster-setting cements, and the industry shifted from cements rich in dicalcium silicate, which harden slowly, to those with higher tricalcium silicate content, which set faster and gained early strength more rapidly. Faster-reacting concrete, however, is harder to work with, and in the absence of effective chemical admixtures, site workers commonly added extra water to improve workability. That practice, well known to be deleterious to long-term strength and durability, offers a logical account of the weaker 1970s concrete. By the 1990s, chemical additives allowed good workability without excess water, and measured strengths rebounded sharply. The researchers caution that the exact causes remain unconfirmed, since detailed mix-design records from the period are unavailable, and that lower compressive strength alone does not by itself establish reduced durability or safety, which also depend on corrosion, carbonation, cracking, geometry and loading history.

For practical use, the study distills its findings into simple linear trendlines relating compressive strength to year of construction, with separate coefficients for each drilling zone and structural element. For tunnel linings, the slope is roughly half a megapascal per year, with slightly lower values for crown concrete than for spring-line concrete. For bridges, the slopes vary from 0.539 megapascals per year for pier elevations down to 0.150 for abutments. These equations are deliberately simple: they are a screening tool, not a substitute for structure-specific testing. Where a preliminary estimate suggests a structure’s concrete may be of concern, detailed destructive and non-destructive investigation should follow. But as a triage instrument for concessionaries responsible for hundreds of assets, the approach promises to redirect scarce funds from blanket testing campaigns toward the structures that genuinely need intervention.

The authors are explicit about the limits of their model. The curves are calibrated specifically on Italian tunnels and bridges built between 1958 and 1995, based solely on compressive strength data, and should not be extrapolated to other regions, concrete technologies or construction periods without local validation. Incorporating variables such as element type, exposure environment, deterioration state and construction practice would yield a more refined multivariable model, and the team notes that further experimental data, including from other European and overseas countries, are needed to strengthen the curves. Still, the framework marks a departure from earlier studies that calibrated strength estimates structure by structure; here, a century-long national database yields general age-dependent trends applicable to an entire population of infrastructure without dedicated calibration for each asset.

The broader message is one of targeted vigilance. As Italy and much of the developed world confronts a generation of concrete infrastructure approaching or exceeding its design life, tools that convert historical data into actionable predictions are becoming essential to sustainable management. The strength-for-age curves suggest that the logical, steady improvement of concrete over the twentieth century has a conspicuous exception, and that the motorway structures of the early 1970s, products of a boom era that prized speed above all, deserve to be at the front of the inspection queue. In a field where the cost of misjudgment was made tragically clear in Genoa, knowing where to look first may prove as valuable as any repair.

Subject of Research: Statistical assessment of concrete compressive strength in aging Italian motorway bridges and tunnels using historical test databases and in-situ core testing

Article Title: Assessing the class of concrete in Italian infrastructures

Article References: Assessing the class of concrete in Italian infrastructures. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: concrete strength, infrastructure assessment, Italian motorways, bridges, tunnel linings, strength-for-age curves, compressive strength, structural safety, aging infrastructure, cement chemistry, core testing, Politecnico di Torino