Utilisation des alliages ferreux dans la construction monumentale du Moyen Age. Etat des lieux de l'avancée des études métallographiques et archéométriques

L'étude métallographique des fers de construction mis en œuvre dans certains grands édifices français des périodes médiévales et modernes apporte de nombreuses informations sur les caractéristiques du matériau utilisé tout comme sur l'histoire de la métallurgie du fer, en même temps qu'elle soulève de nombreuses interrogations. Les résultats accumulés depuis quelques années permettent de mettre en valeur différents points. Tout d'abord, il semble que les grands édifices médiévaux sont pour la plupart renforcés uniquement avec du fer de réduction directe. En région rouennaise, la période de transition identifiée par analyse des inclusions se situe à la fin du XVe siècle, ce qui correspond à l'étude archivistique. Le fer employé, rempli d'inclusions, souvent phosphoreux, et n'ayant pas fait l'objet de traitement spécifique est un matériau standard pour la période médiévale ; de rares exceptions sont cependant à noter. Sont également évoquées les questions relatives aux différentes étapes du travail de forge, et en premier lieu celle de l'utilisation du marteau hydraulique pour faciliter la phase d'épuration de la loupe, opération longue et épuisante d'après les résultats d'archéologie expérimentale. Pour finir, la fréquente homogénéité de certains rapports dans les inclusions semble montrer que les hétérogénéités de la matrice, notamment la diversité des degrés de carburation, sont liées à l'hétérogénéité initiale de la loupe de fer, et non la marque d'une récupération par corroyage.The metallographic study of construction iron used in churches and other buildings during the Middle Ages and the modern period can provide a great deal of information on the characteristics of this material as well as on the history of iron and steel metalwor-king. At the same time the study of construction iron raises many questions. This paper discusses results accumulated from certain French monuments, these include: the cathedrals of Amiens and Rouen, the church Saint-Ouen of Rouen, the basilica of Saint-Denis, the Palais des Papes in Avignon and the dungeon of the Château de Vincennes. The analysis of the composition of slag inclusions entrapped in the metallic matrix allows differentiating iron of the direct process from iron that comes from finery. It seems that all these medieval buildings were essentially reinforced with bars of bloomery iron, and thus didn't really take advantage of the introduction of the blast furnace. In Normandy, where several buildings were studied, the results are compatible with that of the archives: a transition between the direct and indirect processes during the end of the 15thcentury. The metallographic analyses show that in general the iron used in those monuments is very heterogeneous, with no peculiar distribution of the iron and steel parts, numerous slag inclusions, and often a high content of phosphorus. Moreover, no specific forge treatment was observed. This material seems to be a random quality iron for these periods, but whose quality was always fairly sufficient according to its section and role. Certain bars of the Palais des Papes are the only exception to this picture so far. Questions related to the different stages of the smith's work at the forge are then evoked. Results from forge experiments, specifically on the cleansing operation that occurs between the reduction and the forming of the object, have been studied. Two phases have been identified in this stage: first compacting the bloom with a wooden hammer to reduce its porosity, second expelling the slag inclusions that arc still entrapped in the metal by folding the piece of metal and hammering it with a steel hammer. The experiment shows that this phase is particularly long and tiring: more than 4 hours to get an ingot out of a 14 kg bloom. Therefore it brings out the question of the use of hydraulic hammers to help the smith in the second phase of the cleansing. Finally, the analysis of the composition of the slag inclusions reveals that, in more than one case out of two, the ratios of certain elements (Al, Si, Ca) are stable in the whole sample. These elements arc a sort of chemical signature of the object resulting from the reduction. Therefore, it seems that all the heterogeneities seen above arc more likely to be linked to the original heterogeneity of the bloom and not the mark of recuperation by welding pieces of metal from different origins. Therefore brand new metal seems to take an important place in the supply of these monumental buildings. It is also sometimes possible to determine, through the slag inclusion analysis, that sand additions were made during the forging phase