Вы находитесь на странице: 1из 8

Arches in masonry structures: Viollet-le-Ducs rationalist theories

I. Tarro
Structural Department, E. T. S of Architecture, Technical University of Madrid, Madrid, Spain

ABSTRACT: In the middle of the 19th century Viollet-le-Duc began the restoration of the church of Vzelay. Its collapsed vaults, its deformed flying buttresses, its leaning walls and, above all, the enormous distortions of the transverse arches in the vaults of the central nave, were the starting point of his theory about the structural behaviour of medieval constructions. His rational approach was unanimously accepted by most architects, archaeologists and historians of the time, and it was not until fifty years after his intervention in Vzelay when his arguments turned into the object of harsh criticism. Viollet-le-Duc did not add any new idea concerning arch bridges to his theory. On the contrary, his rationalist arguments, used for arches and vaults of medieval buildings, were also used to explain the behaviour of arch bridges in his scarce studies about these constructions. In this paper both Viollet-le-Ducs rationalist theories and his application to arch bridges will be carefully analysed.

INTRODUCTION

The publication of the Dictionnaire raisonn de larchitecture franaise (Viollet-le-Duc 1854-1868) in the second half of the 19th century prompted a turning point in the course of the studies of medieval construction. In this master piece, Viollet-le-Duc proposes a novel interpretation of this kind of architecture, where the technical, functional and constructive aspects become the main topic of all the analysis. This new structural approach, based on his experience as an architect, and above all as a restorer, has given rise to his so-called rationalist or classical theories. Although his theories about masonry structures are basically focused on buildings, on the analysis of their materials, their bond, their geometry, their structural behaviour, etc., they can be perfectly extrapolated to other masonry constructions. That is precisely what Viollet-le-Duc does when he studies medieval bridges. 2 THE ORIGIN OF VIOLLET-LE-DUCS RATIONALIST THEORIES

Around 860 Girard de Roussillon, duke of Lyon, founded a monastery in the valley of the Cure, at the foot of the hill of Vzelay. After being destroyed by the Normans, the building was reconstructed at the end of the 9th century on the top of the hill, just where it stands nowadays. During the 11th and the 12th century the church was at the height of its power and fortune. A rumour spreading that Madeleines relics where housed inside it, turned this monument into one of the most popular places of the Christian world. The extraordinary development of pilgrimage required an extension of this temple that became one of the biggest French churches of the time (120 m long). However, from 1279 and coinciding with the moving of the saint skeleton, the abbey began to decline. Plundered by the Huguenots, badly restored by Abbot Erard de Rochefort and totally unattended during the 17th and the 18th centuries, the monument was decaying slowly with the passage of time. The night of the 21st of October 1819, the tower was struck by a lighting, and for the sixth time, the old sanctuary was burned. When Prosper Merime, Historical Monuments inspector, visited the abbey of the Madeleine of Vzelay in 1834, he discovered the building in a pitiful state of conservation (see Fig.1). The

I. Tarro

339

alarm in view of the risk of an imminent collapse was such that if the assistance to the Madeleine is delayed much longer, it will soon be necessary to make the decision to demolish it in order to avoid accidents (Merime P., 1835). They were afraid that even a slight shaking caused by the consolidation attempts could lead to a looming catastrophe.

(a)

(b)

Figure 1 : (a) State of the facade of the church of Vzelay before the restoration, (b) View of the inside of the church.

On February 1840, on condition that only those parts which could not be preserved should be reconstructed, Viollet-le-Duc was entrusted the restoration of the church. He was only a twenty-six year-old architect, when he left to the small village of Vzelay in the French Burgundy and found the moribund and poorly built construction. In his Rapport sur letat actuel dancienne Eglise de la Madeleine Vzelay, submitted to the Commission des Monuments Historiques on March 1840, he described the deplorable state of the basilica: buttresses not offering a sufficient support to the thrusts of the vaults, walls leaning outwards (over 25 cm), transverse arches completely distorted, and cracked and collapsed vaults damaged by water infiltrations. He claimed that the degradation was so big that if urgent repairs were not undertaken the nave and the choir would collapse. He also suggested shoring temporarily up the walls and placing centerings under the flying buttresses, the transverse arches and the vaults (Jokilehto J., 1986). Viollet-le-Duc began his restoration in Vzelay with the reconstruction of the extremely ruined highest vaults and the demolition and rebuilt of walls and buttresses. He shored up the walls of the apse and rebuilt the deformed flying buttresses of the choir. After the collapse of a vault in 1843 he involved himself in the consolidation of the narthex, the facade and the tower. The vaults in the collateral naves were lightened and the roof of the ambulatory was redesigned. When in 1856 all the work was finished, Viollet-le-Duc had managed to give back to the building, the strength and the beauty of its distant youth (Pore 1909). Throughout the sixteen years he spent on his intervention in the Madeleine of Vzelay, Viollet-le-Duc observed, experimented and analysed in detail this splendid specimen of medieval architecture. His conclusions from the distortions of the prominent transverse arches, from the cracking on the vaults and from the behaviour of the buttress system, set up the foundations of his rationalist theories about construction, geometry and structural behaviour in masonry constructions of the Middle Age. These theories were published in a series of articles in the Annales Archologiques (Viollet-le-Duc 1844-1847) and subsequently gathered together in his Dictionnaire raisonn de larchitecture franaise.

340 3

ARCH10 6th International Conference on Arch Bridges THE RATIONALISM APPLIED TO MASONRY ARCH BRIDGES

According to Viollet-le-Duc there was a universal principle in medieval architecture: the rationalism. Romanesque builders and, primarily, gothic masters used this logic in all their creations. Fruit of this rationality, each vault was associated with a series of arches; the number and the position of these arches conditioned the design of springers, whom also played a basic role in the shape and the dimension of abacus and capitals; and finally, the number of capitals set the location of supports. Thus, constructions in the Middle Age became authentic rationalist pieces, where all the building could be perfectly deduced, in a natural way, from the vaulting system. With his exhaustive descriptions about the design and the geometry of the ribs in the springing of the vault, this French architect tried to demonstrate that even the most insignificant detail was not trusted to chance. Viollet-le-Ducs work is riddled with all kind of explanations and technical arguments for medieval buildings. However when he deals with the bridges of the same time, he is more descriptive than when he is referring to the arches, the vaults or any other element of religious architecture. Nevertheless, it is easy to realize that his theories of arches behaviour in vaulted constructions are exactly the same as the ones he applies to masonry arch bridges. 3.1 Geometry in medieval arches

Viollet-le-Duc knew of the concept of line of thrust (although he did not apply it to his analysis of medieval architecture) and he also knew that this line must be contained between the intrados and the extrados of the arch, for the equilibrium to be possible. As a result, the flattener the arch was, the bigger its thrust would be. Medieval builders were unaware of this theory and they based all his constructions in their intuition, their experience, and the application of rules. First Romanesque groin vaults, provided with round arches, looked similar to a dome because of the higher rise of the diagonal arches. The great distortions suffered by these arches (Viollet-le-Duc takes as reference the transverse arches in Vzelay) was one of the factors that motivated the adoption of point arches in the Middle Age (see Fig.2). With this new shape, the thrusts were considerably diminished, not only due to the path that they described, but also for the amount of weight of the webs that is vertically transferred to the abutments without developing any thrust.

(a)

(b)

Figure 2 : (a) Distortions in the church of Vzelay, (b) Design of a point arch. (Viollet-le-Duc 1854-68)

The use of point arches in the rib vaults of the High Middle Age had, besides this decrease of the trusts, another obvious purpose: to avoiding domed shapes in vaults. To achieve this aim, gothic constructors realized that both transverse and longitudinal arches should have similar rise to diagonal ones. Bearing in mind that the span of the first two arches is quite smaller than the last one, the only way to reach the desired proportions was by utilizing stilted or point arches (see Fig.3). In a nutshell, according to the father of the rationalist theory, point arches began to be used in medieval building for two main reasons: thrust reduction and aesthetical taste. As arches in vaulted buildings, the shape of the arches in bridges exerts a strong influence over the thrust transferred to the piles. Viollet-le-Duc was aware of this and praised the three-arch-circled shape of the arches in the bridge of Saint-Bzenet (Avignon), see Fig.4. The thrusts with this shape are smaller than the ones with a round arch. Most of the masonry bridges

I. Tarro
th

341

built during the 12 century have semi-circular arches; therefore, the use of that new geometry in the above mentioned bridge (built between 1178 and 1188) was a great advance in the field.

(a)

(b)

(c)

Figure 3 : Geometry of romanesque : (a) and gothic (b and c) vault (Viollet-le-Duc 1854-1868)

Figure 4 : Bridge of Saint-Bzenet, Avignon (Viollet-le-Duc 1854-1868)

Point arches in bridges were a constant from the 13th century (see Fig.5). This fact answered both to an aesthetical matter according to the taste of the time, and to functional reason: the more pointed, the less thrust was transferred to the piers. Referring to the last reason, Viollet-le-Duc stated that as the construction of a medieval bridge generally lasted for ten and twenty years, piles would usually receive thrusts coming only from one side. In this situation, if the thrusts were minimized, the piers would stand and, therefore, be equilibrated more easily. Thus, the advantage of thrusts in point arches is extremely useful in these cases.

Figure 5 : Bridge of Calandre, Cahors, built in the 13th century (Viollet-le-Duc 1854-1868)

Nevertheless, the reasons for utilizing point arches in masonry bridges were not always the same as in vaults in buildings. The idea of increasing the rise of an arch by using pointed shapes has no sense in bridges; first, because these constructions did not require excessive heights (just to be enough for a possible river level rise) and, second, because although certain heights were required, the shape of the arches was not so conditioned, hence their span could be easily altered. However, any reason for the introduction of point arches in masonry building or bridges brought a common consequence for both constructions: the reduction of the thrusts transferred to the abutments.

342

ARCH10 6th International Conference on Arch Bridges

3.2 Construction and structural behaviour in masonry arches: the elasticity and the independence The structural behaviour of medieval monuments is determined by their constructive system. In this sense, Viollet-le-Duc stated that, as result of the continuous settlements and distortions suffered by Romanesque and Gothic masonry works (made of rubble walls with mortar infill, see Fig.6a), the masters in the Middle Age built according to the principle of the elasticity. It is important to remark that this concept of elasticity is not the one used nowadays in the modern theory of structures. This seems to be precisely what Pol Abraham did not understand when he criticised the rationalist architect for attributing elastic properties to masonry structures (Abraham 1934). In Viollet-le-Ducs view, the elasticity is not the property of getting the original shape back after a distortion, but the ability to keep a distorted shape without collapsing. In this way, medieval vaults and arches are elastic because they are able to adapt themselves to movements, by distorting their shape and creating hinges (see Fig.6b). This elasticity increases with the reduction of the size of the voussoirs and the distance between the intrados and the extrados, and it is only achievable when the independence between all the elements of the structure is possible. Referring to Vzelay, Viollet-le-Duc wrote builders had already obtained, thanks to the set of independent and resistant transverse and longitudinal arches for each vault, an elastic sketelon over which, if a settlement occurred, vaults could move independently (Viollet-le-Duc 1996). According to this well-known French architect, the elasticity and the independence between elements were two notions that went hand in hand in medieval constructions.

(a)

(b)

Figure 6 : Romanesque masonry wall : (a) Typical distortion in a round arch, (b) (Viollet-le-Duc 1854-68)

Going on these two concepts, for the rationalist master, vaults were made by a series of independent panels resting on a skeleton of arches (see Fig.7). This skeleton worked as permanent and elastic centering that was independent from the rest of the construction. He asserted that gothic builders understood, better than romanesque builders, the independence in masonry structures, and they demonstrated it by adopting a system where all the forces were active forces: at the end of the 12th century buildings consisted of a skeleton stabilised by the combination of an oblique or vertical load equilibrating the thrusts, and of an enclosure, a covering over the skeleton. All the building has its skeleton; a stone skeleton independent of the covering. It is rigid or flexible, depending on the necessities and the place; it yields or resists; it seems to be alive, as it contains opposite forces, and its immobility is due to the equilibrium of the forces that are not passive at all, but active (Viollet-le-Duc 1996). Viollet-le-Duc was very explicit and precise when he described the structural behaviour of medieval buildings. Nevertheless, in his descriptions of bridges, he was not so meticulous (he did not specify how these great masses of masonry were erected). At this respect, the independence of the elements of a construction is a topic that constantly concerned this French architect, although he did not always study it with the same depth. For example, when he studied the independence in a bridge arch, he only differentiates between juxtaposed and not-juxtaposed voussoirs, giving some descriptive examples, but without explaining the consequences of the use of one solution or the other.

I. Tarro

343

Figure 7: Stone skeleton and webs in barrel and rib vault (Viollet-le-Duc 1854-1868)

In his article bridge of the Dictionnaire, Viollet-le-Duc retrieves the concepts of elasticity and independence in masonry constructions with the purpose of explaining the structural behavior of bridges and their arches. He uses these concepts in the same way as he had done before to state that arches and vaults in masonry buildings had to be elastic in order to guarantee their stability. In other words, his explanations about the behavior of masonry arches and vaults are also applicable to medieval bridges, and the analogy between buildings and bridges, from a structural point of view, is clearly manifest in his publications. Viollet-le-Duc does not include new arguments in his rationalist theory when he analyses the behavior of masonry bridges, on the contrary, he just extrapolates to these constructions his previous ideas, conceived during his work in Vzelay. In this way, he considers that the bridge arches and the spandrels are independent elements that work separately, just like the transverse arches of medieval vaults and the webs supported by them. The arches of Poitou bridges are examples of authentic transverse arches, which, in addition to being a more inexpensive alternative, since they require less material for the construction of the bridge, they produce a lighter structure with less thrusts (see Fig.8a).

(a)

(b)

Figure 8 : Details of the bridge of Poitou : (a) and Saint-Bzenet, (b) (Viollet-le-Duc 1854-1868)

From Viollet-le-Ducs point of view, the elasticity of arch bridges means that they are extradosss, i.e., their extrados is carved on a regular basis and concentric with the intrados. When this happens, arches can move independently from the spandrels and the vibrations caused by the loads exerted on the bridge (typically live loads) only affect the spandrels. These vibrations do not affect the arches; they are transferred directly to the piers without disturbing the rest of the construction. The engineer Jean-Rodolphe Perronet (1708-1794) used to clamp the voussoirs of his bridges with iron pieces with the aim of providing additional rigidity to the structure and allowing all its elements working as a single piece. According to Viollet-le-Duc, this practice was an error that proved that modern constructors did not understand at all the principles of masonry construction. For him, the connection between the elements of the constructions, far from being an advantage, would turn with time into an active destructive agent of the masonry (Viollet-le-Duc 1854-1868). Making a structure more rigid is making it more vulnerable, as their elements need to be independent in order to move one with respect to the other. In order to achieve the

344

ARCH10 6th International Conference on Arch Bridges

independence of the structural elements of a bridge, the joints between the arch pieces and the rest of the bridge should not be filled with mortar (see Fig.8b). However, this physical independence does not prevent the different parts of the bridge from working together. Masonry structures are distortional structures (elastic, according to Viollet-le-Duc) and, thus, able to move, crack and hinge, in search of a possible equilibrium state. This adaptation ability is precisely what made possible that masonry constructions had survived until these days. 4 VIOLLET-LE-DUCS RATIONALIST APPROACH IN THE CONTEST OF GOTHIC STRUCTURAL THEORIES As stated above, when Viollet-le-Duc analyses arch bridges behaviour, he bases all his reasoning on the rationalist theories he developed during the restoration in Vzelay. Throughout many years, his rational point of view, based on his knowledge of structural mechanics, was accepted by most scholars in the field, not only in France but also in other countries. Probably the most faithfull and distinguished follower of the rationalist theory was Auguste Choisy (1841-1909). His arguments over gothic constructions, gathered together under the title of Histoire de larchitecture (Choisy 1899), had huge popularity and contributed to the spreading and credibility of his predecessors approach. Nevertheless, since Viollet-le-Ducs death, the rationalist theories turned into the object of criticism from numerous researchers in the field of gothic architecture. The first one in showing his disagreement with the classical theories was Anthyme de Saint-Paul in 1880. In his long essay, Viollet-le-Duc et son systme archologique, he tried to proof the lack of logic and rationalism of gothic constructions, and therefore the falseness of the theories of those who supported the primacy of the structural role in some elements of this architecture. After him, a large number of authors followed this idea calling into question the validity of the classical or rationalist approach (Brutails, Enlart, etc.). This was the beginning of a lively and intense debate about gothic structures. Supporters of the constructive and rational logic in medieval buildings and its detractors, who defended a decorative and aesthetic function in those elements to which the rationalist theory attributed a fundamental structural role, were confronted during more than a hundred years. The decade of the 1920s was one of the most prolific periods in the debate. In 1920 Roger Gilman carried out an exhaustive study about the ruins caused by the bombings of the First World War in Reims and Soissons cathedrals. The interest of this investigation lies, not in his conclusions, but in the analysis method that he uses, which is based, for the first time, on the observation and the analysis of damaged and partly-destructed constructions. This new method has been subsequently reused by some other authors in order to explain how masonry constructions work (T. Balbs). At the end of the 1920s the French engineer Victor Sabouret published what could be considered as one of the first big criticism to Viollet-le-Ducs rationalist ideas. Focusing on vaults behaviour he attributed a non-structural role to the arches (Sabouret 1928). Nevertheless, the most devastating attack to the principles of the classical theory was due to the French architect Pol Abraham. In his book, Viollet-le-Duc et le rationalisme medieval (Abraham 1934), Viollet-le-Duc was subjected to the merciless criticism of this author. According to Abraham, the webs are self-supported elements and the ribs have no structural usefulness (they are only decorative elements) and therefore the rationalist premises are wrong, unwarranted, anachronistic, contradictory, and based on a priori ideas. His explanations, which were often incorrect, are due to the eagerness to delegitimize Viollet-le-Ducs hypothesis. Pol Abrahams ideas had an enormous influence over the following generations (Aubert, Masson, Focillon) and the debate continued with great strength over the decade of the 1930s. Unfortunately, this proliferation of studies in the field of gothic structures was truncated by the outbreak of the Second World War. The little interest on the subject, from the second half of the 20th century, consigned this hundred-year-old debate to the oblivion. Thanks to the contribution of the engineer Jacques Heyman the debate reached its end in the middle of the 1960s. His studies about gothic construction led him to the theory of the limit analysis applied to masonry structures. This new theory lays the foundations for a scientific

I. Tarro

345

analysis on gothic structural behaviour, providing us with a theoretical framework from where it is possible to assess the accuracy and the suitability of all the discussions of the debate. As explained previously, most of the references to Viollet-le-Duc, against and in favour, were specifically focused on the structural behaviour of the vaults of medieval buildings. He also used his theory to explain the behaviour of medieval bridges, in the same way as he did with medieval buildings. Thus, although there are no explicit allusions to masonry bridges coming from his supporters and opponents, his ideas about arch bridges, described in section 3, would be necessary subjected to the same criticism and support. In conclusion if a light wants to be shed on masonry bridges behaviour, these constructions must be analysed in the framework of Heymans theory. 5 CONCLUSIONS

In Viollet-le-Ducs opinion, all the constructions in the Middle Age were ruled by a rationalist principle, according to which, in medieval architecture nothing was trusted to chance. In his work, he does not make much mention to arch bridges. Nevertheless, although his arguments are not so detailed as when he refers to buildings, his approach is the same: the technical, constructive and structural point of view that revolutionized the researches on medieval architecture. Referring to the geometry of the arches, he stated that the introduction of the point arch in medieval constructions contributed to minimize the thrusts transferred to the abutments at the same time that an aesthetic shape was achieved. His explanations about masonry structural behaviour are based on what he called elasticity: masonry structures are elastic, i.e., they become distorted by moving and cracking without collapsing. This property is only possible if all the elements of the constructions are independent. All his ideas had a great influence over the subsequent generations and provoked a long debate on structures behaviour that ended, after more than a hundred years, with the contribution of Heyman to this field. Even though most of the arguments were focused on vaults and buildings behaviour, Viollet-le-Ducs ideas about bridges were also the object of all the discussions as they are an extrapolation of his rationalist theory. REFERENCES
Abraham P., 1934. Viollet-le-Duc et le rationalisme mdival. Paris: Vincnt, Fral et Cie. Choisy A., 1899. Histoire de l'architecture. Paris: Gauthier - Villars. Gilman R,. 1920. Las teoras de la arquitectura gtica y el efecto de los bombardeos en Reims y Soissons. Arquitectura, 3, p.216-222; 237-254. Heyman J., 1966. The Stone Skeleton. International Journal of Solids and Structures, 2, p. 249-279. Jokilehto J., 1986. A history of architectural conservation: The contribution of English, French, German and Italian towards an international approach to the conservations of cultural property. Ph.D. diss, University of York. Merime P., 1835. Notes dun voyage dans le Midi de la France. Paris: Librairie de Fournier. Pore C., 1909. L'abbaye de Vzelay. Paris: Henri Laurens. Sabouret V., 1928. Les votes d'artes nervures: Rle simplement dcoratif des nervures. Le Gnie Civil, 92, p.205-209. Saint-Paul A., 1880-1881. Viollet-le-Duc et son systme archologique. Bulletin Monumental, 46, p.405-463, 715-775; 47, p.5-54, 187-235, 349-417, 445-502. Viollet-le-Duc E.E., 1844-1847. De la construction des difices religieux en France. Annales Archologiques, 1, p.179-186; 2, p.78-85, 143-150, 336-349; 4, p.266-283; 6, p.194-205, 247-255. Viollet-le-Duc E.E., 1854-1868. Dictionnaire raisonn de l'architecture franaise du XI au XVI sicle. Paris: A. Morel. Viollet-le-Duc E.E., 1996. La construccin medieval. El artculo Construccin del Dictionnaire raisonn de l'architecture franaise du XI au XVI sicle. Ed. by E. Rabasa and S. Huerta. Madrid: Instituto Juan de Herrera/ CEHOPU.

Вам также может понравиться