The fire that destroyed large sections of the iconic cathedral Notre-Dame de Paris last April was a national tragedy. Now, months on, scientists with the French national research organization CNRS are embarking on a multimillion-euro effort to study the 850-year-old building and its materials with the goal of illuminating how it was constructed. With unprecedented access to the cathedral’s fabric — including timber, metalwork and the building’s foundations — in the wake of the fire, scientists also hope that their work will arm them with information to help the restoration.
The research could “write a new page in the history of Notre-Dame, because there are currently many grey areas”, says Yves Gallet, a historian of Gothic architecture at the University of Bordeaux-Montaigne, who is in charge of a 30-strong research team investigating the masonry.
Construction of the cathedral, considered one of the finest examples of the French Gothic style, began in the twelfth century. The structure was modified in the Middle Ages and extensively restored in the nineteenth century by the architect Eugène Viollet-Le-Duc. But it has been the subject of surprisingly little scientific research, compared with other Gothic monuments in France and elsewhere, says Martine Regert, a biomolecular archaeologist at the CNRS’s CEPAM centre for the study of historical cultures and environments in Nice, who is one of the Notre-Dame project’s leaders. Many questions remain about the structure, such as which sections are medieval and whether Viollet-Le-Duc reused some of the older materials, says Regert.
The fire on 15 April, possibly caused by an electrical fault, destroyed the cathedral’s roof and spire, and caused part of its vaulted ceiling to collapse. The walls still stand, and the building will eventually be restored — although this is likely to take longer than the ambitious five years initially forecast, and is set to cost hundreds of millions of euros.
But until then, the interior of the building holds piles of debris: fallen stonework, burnt timbers and damaged metal artefacts, all now available for scientific study. The absence of tourists might also make it possible to use radar imaging to probe the foundations, which have been little investigated. Even some parts of the structure that were largely undamaged are now more accessible for inspection, says Philippe Dillmann, a specialist on historical metal artefacts at the CNRS Laboratory for Archaeomaterials and Alteration Forecasting in Gif-sur-Yvette, who is coordinating the project with Regert.
The CNRS project will focus on seven topics: masonry, wood, metalwork, glass, acoustics, digital data collection and anthropology. In all, the effort will involve more than 100 researchers in 25 laboratories and will last for 6 years.
Gallet’s team will study Notre-Dame’s stones to identify the quarries that supplied them and “reconstruct the supply networks and the economy of the site”. Studying the mortar used to bind the stones together could reveal how different compositions were used for the various structural elements — vaulting, walls and flying buttresses. The mortar used lime prepared from sedimentary limestone, which might contain fossil remnants that could reveal where it originated. A better knowledge of the historical materials could inform choices made in restoration, says Gallet.
The team will also analyse weaknesses in the remaining structure caused by the high temperatures of the fire, the fall of masonry and the water used to extinguish the flames. Damage to the stones was exacerbated last July by extreme heat waves in Paris, which “brutally dried” and weakened the masonry, says Gallet. A radar study will determine how solid the foundations are before restorers erect scaffolding in the crossing between the nave and the transept to allow them to dismantle the unstable remnants of the nineteenth-century spire.
And with the help of historians, Gallet’s team hopes to gain a deeper understanding of the structural engineering of Gothic architecture as a whole, and Notre-Dame’s place in that story.
Out of the ashes
Meanwhile, a team of about 50 will focus on Notre-Dame’s famous woodwork — especially the ‘forest’ of timbers in the roof space above the vaults — which has either burnt away or lies charred in the nave. These blackened remains could be tremendously valuable to researchers.
“The burnt structure constitutes a gigantic laboratory for archaeology,” says Alexa Dufraisse, an archaeologist at the National Museum of Natural History in Paris, who will lead the multidisciplinary wood team. The group will include archaeologists, historians, dendrochronologists, biogeochemists, climatologists, carpenters, foresters and engineers specializing in wood mechanics.
“Wood is an extraordinary source of information,” says Regert. Initial observations have confirmed that the ‘forest’ is made of oak, but studies will pinpoint the exact species used and give researchers clues about the techniques and tools of medieval timber construction.
Tree-ring dating of timber beams could reveal the year and location in which the trees were felled, filling in gaps in knowledge about the sequence of construction. “Each tree records within its tissues the environment in which it has grown,” says Dufraisse. This kind of study “could never have been conducted without the destruction of the structure by fire”, she says.
In particular, says Regert, the wood is a climate archive. “Isotopic analyses of oxygen and carbon in the rings make it possible to determine the temperature and rainfall over time,” she says. The trees used in Notre-Dame grew between the eleventh and thirteenth centuries, during a warm period known as the medieval climate optimum, offering a reference period for natural climate warming to compare with anthropogenic warming today. “This period is poorly known because woods of that time is rare,” says Dufraisse.
Metal and masonry
A separate team will investigate the cathedral’s metalwork — in particular that used to support the stone and woodwork. “We want to understand the use of iron armatures in the different construction and restoration phases,” says archaeologist Maxime L’Héritier of the University of Paris 8, who will lead the study. Metal rods, for example, were used to support sections of masonry under tension, and medieval builders sometimes inserted iron chains into the stonework to strengthen it. L’Héritier says that there has never before been a study of changes in the use of iron in cathedral building over such a long period, from the Middle Ages to the nineteenth century.
His team will also study the lead from the roof — much of which was damaged or melted in the fire. The researchers aim to develop a chemical reference data set that records the ratios of lead isotopes and the presence of trace elements in the material, “to understand the evolution of lead quality and supply” — for example, to identify the mines from which the metal came. The group also wants to investigate how much lead was recycled when the roofing was restored in the nineteenth century. These results might also enable researchers to work out how much lead the fire released into the environment — a potential health hazard for the immediate vicinity.
Access all areas?
Collecting and excavating the materials for analysis is challenging. There are three main piles of debris — in the nave, the crossing and the north transept — as well as material still on top of the remaining vaults. But these are currently off-limits to people for safety reasons, Dillmann says — so robots and drones must do all the collecting. Some of this material might ultimately be reused in restoration.
“The first challenge is to collect all wooden elements, regardless of their level of carbonization,” says Dillmann. So far, he says, nearly 1,000 fragments have been collected and labelled — but the work is just beginning. Dufraisse says that this wood won’t be accessible to researchers for at least another three months, because it is currently too contaminated with lead. Researchers will need to calibrate how chemical signatures in the wood have been modified by the high temperatures of the fire. “I know we are going to be faced with technical problems, but I remain confident,” says Dufraisse.
The collection and analysis will need to be documented precisely and thoroughly. Livio de Luca, a specialist in digital mapping of architecture, at the CNRS’s Mixed Research Unit in Marseille will lead a team dedicated to creating a “digital ecosystem” that summarizes both the scientific research and the current and previous states of the cathedral, drawing on the work of scientists, historians, archaeologists, engineers and curators — and perhaps even on old tourist photos of the structure.
“It will be like a ‘digital twin’ of the cathedral, able to evolve as the studies progress,” de Luca says. It will include online models for 3D visualization of the building and its attributes — a kind of Google Earth for Notre-Dame, created from billions of data points, with the history and evolution of the structure superimposed on the spatial map.
As well as deepening our understanding of this monumental building, Regert hopes that the scientific studies will be useful when its ravaged vaults rise again. The results, she says, might “illuminate the choices that society will have to make for the restoration”. She hopes, too, that they could help to prevent such a catastrophic accident from happening again.
Nature 577, 153-154 (2020)