Monument Future. Siegfried Siegesmund
Чтение книги онлайн.
Читать онлайн книгу Monument Future - Siegfried Siegesmund страница 24
Traditional renders in the Vexin Français
Traditional houses in “Vexin Français” can be classified into three groups: big farms for growing of cereals with closed court-yards; workers’ houses and wine growers’ houses. Before the arrival of railway in the area, the price of raw materials forced people to build with local quarry materials: mainly limestones, sandstones in some areas or “meulière” (millstone). Ashlars and “meulière”, prestigious materials, were apparent but rubble stones were covered by renders.
Nowadays, many people think that plaster of Paris is a weak material, soluble in water (~2 g/l at 20 °C) and very sensitive to environmental conditions. Gypsum has been used in façades since antiquity and it has been usually employed in the last century in very different climate areas (Spain, France, Germany, Czech Republic, etc.). Some of these renders are well preserved. Blottas (1839) showed that gypsum at Montmartre (Paris, France) is a sulphurous limestone (“chaux sulfatée calcarifère”) composed of gypsum with more than 12 % of calcite. This mineralogical composition, (Blottas, 1839; Debauve 1884; Flavien 1887), does not correspond to the amount of calcite actually found in Parisian gypsum, varying from 2 % to 5 %. In traditional Plaster of Paris we find overcooked, undercooked and uncooked gypsum due to the inhomogeneity of the temperature distribution inside the kiln. It can be thought that the high durability of gypsum plaster with respect to outdoor environmental conditions comes from the presence of other minerals as clays (Sanz Arrauz and Domínguez, 2009).
Different kinds of renders have been used in the Paris area. Some of them are mixtures of plaster of Paris and aerial lime, around 12 % (Thénard, 1834; Le Dantec, 2016). In some cases the amount of lime in gypsum renders can be explained by a batching of plaster with lime water. Some gypsum-free mortars have been employed in the region, as mortars or as renders (Toussaint, 1841). Lime mortars are more difficult to apply, they harden slower than gypsum and it is less easy to achieve a flat surface. Rugosity of lime mortars is higher than that of gypsum renders due to the present aggregates and for this reason they are less used as finish renders in this region.
Materials and Methods
In this work we studied renders sampled in old façades and new renders formulations in order to compare their physical properties, composition and microstructure. New plasters are used in the rehabilitation of vernacular buildings but also in the restoration of buildings and monuments in historical cities.
Twelve fragments of façades from the Vexin Français area and about ten render formulations supplied by Plâtrière Vieujot were studied. Some façade samples come from the Tiffanie Le Dantec collection (TLD) (Le Dantec, 2019), while others were collected from the sites during this study (CN). They were taken before rehabilitation campaigns or from walls where the renders were about to be taken off. Some of the samples presented two different layers that were characterized separately. It is very difficult to date these renders by documents as most of the studied buildings are private houses or farms without any record on ancient restoration.
TLD collection consists in samples from:
— Théméricourt (THM, XVII century): A façade with two different states of weathering and different colour layers.
— Condécourt (CON, Château de Villette, 1663–1669): Façade with two layers of render, a grey layer and a white one.
— Hédouville (HED, XIX century): Two samples, one from the front façade with a grey upper layer and a white inner layer, and one from the rear façade with only one white layer.
— Fontenay Saint Père (FStP, XVIII century): White façade.
— Arthies (ART, probably XX and XIX centuries respectively): Two samples from one white farm façade, called “Moellons” et “Charbons”.
67CN collection
— Fontenay Saint Père (FStP): White façade, quite degraded
— Saint Clair sur Epte (StC): Two façades with lime render with large amount of aggregates.
— Genainville (GEN): Small part of a white render, almost completely disappeared.
— samples from non-well-maintained enclosure walls at Saint Gervais and Ambleville (W1, W2, W3) We observed façade samples under polarized light and digital incident light microscopy.
Formulations of new render samples of “Plâtrerie Viejot” have also been fabricated according to the experience of this company specialized in outdoor and indoor plaster renders. They reproduce in industrial kilns the traditional cooking of gypsum in “four culée” kiln type (XVIII and XIX centuries).
Water has been added manually to the powder until desired workability was achieved. This procedure corresponds to what is done during building works depending on the experience of the plasterer. For these renders only general results are presented as manufacturer consider the data confidential. Formulations consist in gypsum with different amount of lime but also with marble powder and charcoal.
Mineralogical compositions were determined by thermo-gravimetric analysis (TGA) and X-Ray powder diffraction (XRD). Porosity and apparent density were measured by the triple weight method (NF EN 1936); water vapour sorption/desorption tests were determined following the procedure presented in standard NF EN ISO 12571 with a Vsorp Plus apparatus. Water vapour permeability was measured by the standard NF EN 15803, method of wetcup. Compressive and flexural strength were measured following NF EN 13892-2 standard.
Most of new plaster samples have been measured about one week after their preparation, and some of them have been re-measured 30 days later. All samples were dried at 60 °C during 48 hours. Not all properties could be measured for all of the samples due to the limited size of historical samples and to the preparation procedure of new ones.
Results
Historic renders
The results cannot be considered representative of all the renders from the Vexin area because mainly gypsum renders were considered. The percentage of gypsum vs. lime renders in the area is unknown. Having survived weathering, sampled renders represent the most durable renders of XIX century. Experimental results show that samples from historic façades have a relatively uniform mineralogical composition (Table 1), close to formulations proposed nowadays. However, their physical properties are quite different, probably due to changes produced with time.
According to TGA analyses, in most of the historic samples, except StC, the amount of gypsum ranges from 80 to 93 % and the amount of calcite from 1 to 8 %. The latter may either be an impurity of the natural gypsum, indicate high temperature calcination, or correspond to the addition of lime or calcite aggregate to the plaster. The amount of non-detected minerals by TGA varies from 2 to 16 %. In order to identify this unspecified portion, XRD analyses were done. Quartz but also anhydrite, 68halite and aragonite were found. In StC samples, only calcite was found with over 80 % of unidentified minerals by TGA, which indicates a lime render with a large amount of aggregates.
Table 1: Thermo gravimetric analysis