Structural Analysis of Historical Constructions - Modena, Lourenço & Roca (eds) © 2005 Taylor & Francis Group, London, ISBN 04 15363799

Study of Place Stanislas coloured pavement for its historical restoration

J.M. Mechling Université Hel1l'i Poincaré - Nancy 1, Dpt Génie Civil - IUT de Nancy-Brabois, Villers-Iés-Nancy, Fran ce R. Elter INRAp, , Fi'ance

ABSTRACT: Listed as a World Patrimony by UNESCO, the Place Stanislas (Stanislas Square) in Nancy () was built in 1755. The ground ofthe Square will soon undergo a complete remodelling which aims at restoring it to its original glory. To prepare the project, a historical and geological study was carried. The research into the archives helped us to discover some unpublished documents. They present the initial state ofthe ground and appearance ofthe Square, previously unknown. They gave some valuable indications about the cobblestones (red and black limestones, layout, geometric design ... etc.) and information about the approximate cobblestone quarries. A geological study was necessary to define the nature and origin of the black cobblestones but doubt remains about the real coloured aspect of the red ones. Some chromatic analyses indicate that ali the different red rocks likely to have been used relatively close to the same ochre brown colour.

INTRODUCTlON but has since been replaced by a statue of the Polish King and last Duke of . 1.1 S/anislas Square origin Listed as a World Patrimony by the UNESCO, Stanis­ 1.2 Res/oration goal las Square was built from 1752 to 1755 by Stanislas Leszczynski, former king of Poland (1677- 1766). To celebrate the 250th anniversary ofthe Square'scom­ It is located south of the Carritire Square between pletion, the urban council ofNancy and its agglomer­ the medieval town and the seventeenth century new ation are undertaking a complete remodelling of lhe town (Pfister 1974). The Square was named Place Square which aims at restoring the original ground Royale until 1792 (), to become and the vicinity ofthe Square. The current weight con­ Place Stanislas in 183 I. The Square used to feature the straints used for a public square have to be respected statue of King Louis XV ofFrance, Stanislas' step-son (pedestrian space with some occasional heavy weight

Figure I. Actual aspect ofthe Stanislas Square (photography: Town ofNancy).

265 vehic\es passages for special events that demand bleachers or stands, tents, etc.). The Chief Architect at Historical Monuments is supervising the project. A study of the archives has been conducted to prepare the ground restoring project. The aim was to discover documents about the original pavement. Search has to be conducted by a geological study to detail the different aspects and colours of the pavement mentioned in the archives. Figure 2. A view of the Stanislas Square sourced from lhe painting ofPange. (Photography: Town ofNancy).

2 NEW HlSTORlCAL DATA sizes had to be of approximately 0.25 to 0.29 m in length, 0.20 to 0.26 m in width and 0.26 to 0.29 m in Until recently, the official known documents of the high. They couldn't be sensitive to frost. Another doc­ Stanislas Square didn't indicate the use of colour nor ument specifies the use ofa good quality rock without geometric design ofthe pavement. In May 2003, orig­ imperfections particularly the c\ay beddings. inai documents were discovered in the Public Record The cobblestones were carried to the bui lder's yard Office ofNancy (Archives 1755). One ofthese docu­ by the pavers (Archives 1755). Thereafter, the cob­ ments, a quantity survey conducted by the architect blestones were placed on a sand or gravei levei with to pay the pavers reveals the component and the a disposition that allowed for crossed joints. The exact geometric designs of the Square grounds at the pavement was then covered with sand. time of its inauguration. Some contracts agreed with the pavers and the quarrymen gave information con­ 2.4 Remarks abotll this new documentation cerning the origin, the quality and the sizes of the cobblestones. Surprisingly, the documents found in the archives gave new information about the two colours of the pave­ ment and their geometric designs totally ignored by 2. 1 Geometric design ofthe square the official iconography. The pavement was essentially red in colour, underlined Only one painting, which can be seen in the castle with black cobblestones lines. The black cobblestones of Pange (sixteen kilometres to the north of Nancy), were located around the base of Louis XV's statue shows how the cobblestones were installed (Fig. 2) . and along the square lateral gutters. They represented Red and black cobblestones are visible as the black the four diagonais between the corners and the central diagonais that leading to the corners to the statue of statue (Fig. 2). Louis XV This pavement was preserved as far as the French Revolution. A new quantity survey from May 30th 1789 indicated that refection works were neces­ 2.2 Quarries location sary (Archives 1789). They were eventually stopped by Other documents gave a wide range of information the Revolution. The monarchy marks as the statue di s­ about the work order from the quarry to the grounds. appeared in 1792. The first geometric pavement design The stones were ali extracted from less than ten kilo­ to highlight the statue were changed and disappeared metres from the Square. They carne from different from the collective memory ofthe inhabitants. local towns, Art-sur-Meurthe for the black cobble­ stones, , Malzéville, Pixérécourt and Vandoeuvre for the red ones. Today, these sites are too 3 FIRST GEOLOGICAL ARGUMENTATION vast and contain different rocks than the past black and red cobblestones. Special geological investigations 3.1 Nancy geological context will be necessary (refer to below). From a geological point of view, Nancy is located on the eastern fringe of the sedimentary basin of Paris. 2.3 Cobblestones and characteristics ofthe The town and its vicinity grow in a c\ayey dip, sur­ builder s yard mounted by a calcareous table-Iand and its residual hill, respectively to the west and north-east. A river The cobblestones were sized on the quarrying place. lhe Meurthe runs inside the dip not far from Nancy's As shown in an agreement between the architect and city centre. Local stratigraphy (dated from the Juras­ the local paver Jean Maizieres on July 15th 1754 sic period) also shows two blocks. At the bottom the and as found in the archives, the work had to be first layer (geological stages Hettangian-Sinemurian, very precise. Indeed, the layout of the pavement pre­ Pliensbaschian, Toarcian) mainly contains c\ays with scribed a selection of cobblestones in the quarries; a few limestones and sandstones on 250 to 270 metres

266 high. The second layer (geological stages Aalenian Table I. Links between the Archives localities and the and Bajocian) is a continuai succession of severa I geological stages. limestones lithofacies on 120 metres deep. The rocks susceptible to have been used for the pavement have to Stages Petrography Colours Local towns be found in some geological strong sedimentary beds White, grey in exclusion oftoo many clayey materiais. These rocks Mid. Bajocian Limestone Low. Bajocian Limestone Ochre brown must correspond to the initial colours criteria. Histori­ Grey Malzéville cal archives could help us to determine more precisely Aalenian Limestone Red, orange Pixérécourt locations of ancient quarries. Brown Vandoeuvre Green, grey Toarcian Marle 3.2 Colours present in locallithostratigraphy Pliensbaschian MarJe/ Dark grey Sandstones Black ArtIMeurthe In this lithostratigraphy, we can effectively find again Hettangian - MarJous- BJue-ish the different colours described in the historical writ­ Sinemurian caJcareous ten documents. The oldest sedimentary beds from Hettangian-Sinemurian, Pliensbashian and Toarcian contain dark lithofacies of grey, blue-ish and black colours. With the exception of a particular limestone black cobblestones whereas Aalenian andlor lower bank of a I or 2 metres thick, named Calcaires ocreux; Bajocian limestones have furnished red materiais. initially black, it can progressively takes a rusty colour We have to complete these previous indications with with the weather alteration. some field observations to confirm the real aspects of Just above, Aalenian geological stage coincides the two sorts of cobblestones. with the regional iron ore well known as Minette Lorraine and extracted during the nineteenth and the first part ofthe twentieth centuries. Its petrographical 4 PETROGRAPHICAL ASPECTS OF THE aspects and colours do vary. In this geological sedi­ COBBLESTONES mentary beel, shades encountered are caused by iron oxides variation and concentration. They can be grey, 4.1 Black cobblestones green, reei, brown, with many sort of rust or orange The case of the black cobblestones is simple beca use shades (Bubenicek 1961). sedimentary bed Calca ires à Gryphées is the only Above, lower Bajocian allows calcareous rocks still one allowing black or dark grey limestone in the containing a few iron oxides. The shades are there­ local statigraphy. Furthermore, they could be no doubt fore systematically ochre brown, rust or even red about the black cobblestones origin as its presents in with a more or less grey colour adjunction. Local Art-sur-Meurthe, was mentioned in the archives. encrinitic limestone, Calca ires à entroques, is occa­ Today, there isn 't any quarry visible in this local area sionally named red rock, Roche rouge, by quarrymen but archives also mentioned a special tenancy rental and geologists. contract for the purpose of extracting those stones. At the top of the local stratigraphic series, colours The original cobblestones sources was about 30 cm from the middle Bajocian are very different from the cubic and covered approximately 7.5% of the Square previous rocks. Shades are typically white-ish like the (about 850 m2 compared to 11300 m2) representing a white oolite sedimentary beel, Oolilhe blanche, or grey­ total volume of 240 m3 . A very small production of ish like coral limestone, Calcaires à polypiers. These quarry was necessary for as the stripping and the waste different specific colours are easily recognizable in the rocks. At the end of the lease and the extraction being road slopes that climb the table-Iand over Nancy. finisheel, farm cultivation replaced the quarry activity. Fifteen kilometres to the west, an important quarry is currently in use in this sedimentary bed by a cement 3.3 Preliminary results company. The local towns of Malzéville, Pixérécourl and Van­ This Calca ires à Gryphées is an alternance of cal­ doeuvre match the Aalenian and Bajocian geolog­ careous and marly banks of 10 cm to 60 cm high. ical stages, whereas Art-sur-Meurthe relates to the Fossils (ammonites, ancient bivalvia like Gryphea Hettangian-Sinemurian. arcuata) can be very abundant in particular places and The comparison of these quarries coming from the the limestone characteristic aspect is generally con­ different towns above named and as mentioned in the stant. The stone texture sub-lithographic is very thin archives, with the first geological argumentation and (Fig. 3) and therefore cobblestones are not very dif­ the colour oftheir extracted rocks is consistent (Tab. I) ficult to size. On the other hanel, these limestones are and allow us to make initial conclusions. Hettangian­ generally a little clayey (about 5%) and also present Sinemurian limestones have been extracted for the some low resistances to the freeze and thaw action.

267 Figure 4. Microscopic view of the ferruginous ooliles in Figure 3. Aspect of the black limestone Ca/caires à Minefte Lorraine (magnification: x 25). Gryphées, note the very thin texture (width of lhe field of view: 14cm).

4.2 Red cobblestones Find the original colour ofthe red cobblestones is more complex than fo r the black stones for two main reasons. At first, a red-ish colour could correspond to three geological calcareous sedimentary beds of geologi­ cal stages Aalenian and lower Bajocian. Secondly the archives from the local towns previously mentioned report these three different red rocks sedimentary bed for systematic extraction. For each one, Minette Lor­ raine, Calca ires sableux and Calca ires à entroques we have to consider, petrography and colours ofthe lime­ stones, aptitude for stone cutting and performances, and the eventual presence of historical quarries. Minette lorraine (Aalenian): In Nancy area, this formation is the only one that can have some really pronounced red colours among others like those previously named in Section 3.2. It's a set of carbonaceous (oolites, ossicles) or siliceous (quartz) Figure 5. Homogeneous aspect of the Ca/caires sab/eux detrital deposit, from 10m thick. There is a high (width ofthe field ofview: IOcm). concentration offerruginous minerais such as iron sul­ phurs (pyrite ... ) and iron oxides (limonite ... ). The quartz grains (between 5 and 40%) and clayey mi n­ rock cementation is made with carbonates (calcite . .. ) erais. Observed through a microscope, most of the and ferruginous compounds. others elements are bioclastic materiais as fragments Around Nancy, this sedimentary bed has been inten­ of echinoderme, serpula and small bivalvia wich are sively and industrially used in underground mines and coloured in ochre brown. Their good wide spread presents a characteristic microfacies of ferruginous among the rock combined to the iron oxides present oolitic limestone (Fig. 4). However these rocks are in the carbonaceous cement, give a stable colour and generally toa clayey and have insufficient mechanical aspect to the rock (Fig. 5). properties to cut cobblestones. Only a few blocks have In a lot of reefs, clays and quartz induce the for­ been used in a wall of a mediaeval castle established mation of thin lamina (I to 2 mrn high), responsible at 40 km south of Nancy in the village of Vaudémont of an important rock breakdown (Castaing 1972). In however nobody is aware of any quarry used for blocks these conditions, a production of cobblestones is not extraction. easy and requires the selection of the most favourable Calcaire sableux (lower Bajocian): banks of stone. This geologicallevel is defined by a thin ochre brown An ancient quarry is known in the local area of li mestone layer that contains a variable quantity of Vandoeuvre and is now used for modem urbanization.

268 close to those two quarries have shown that encrinite usually have two coloured different aspects. The more cornrnon is a grey facies with red orange milimetrik or a couple of centimetres large spots (Fig. 6) caused by a high enrichment of iron oxides (limonite ... ). But this limestone can regularly be more homogeneous with a very thin orange spotty aspect (Fig. 7). In this particu­ lar case, iron oxides are well scattered in carbonaceous cement and porosity from small size. The sedimentary bed gives acceptably good mech­ anical properties but is not really suitable for a cutting. Figure 6. Heterogeneous aspeet of the Ca/eaires à Actually many quarries uses this rock for crushed gran­ entroques. Note the differenee between the bottom part ofthe ulates generally used as seating ofroads. From a nor­ sample (grey limestone with many small blank ossieles) and mative aspect, impact strength (Los Angeles test, NF the large red spot above (width ofthe field ofview: 15 em). P 18-573) or attrition strength (Micro Deval test, NF P 18-572) present some moderate results (LA = 30 and MDE = 40) (Zennir, 1996). However those mechani­ cal properties are better than those of the two previous formations (Minette Lorraine and Ca /caires sableux) that are never used in the actual public works.

4.3 What does the red rock rea//y look like? It 's very difficult indeed impossible to give the real aspect ofthe red rocks as simply named in the archives. In some cases, it can correspond to an unvarying rock aspect with a more or less pronounced ochre brown colour (Calcaires sableux and Calca ires à entroques in several banks). In some other cases, it corresponds to a spotty rock resulting from the combination of a grey­ ish colour with red orange to rusty spots (Ca/caires à entroques). This first aspect is the same than the one represented Figure 7. Relatively homogeneous appearanee of the on the painting kept in the castle of Pange (Sec­ Ca /eaires à entroques presenting a thin spotty aspeet (width tion 2.4). Meanwhile, many ancient quarries would ofthe field ofview: IOem). principally furnish the second aspect. The uncertainty persists. It may be possible that other ones existed and progres­ In our historical remodelling, we can only define the sively disappeared for the same reason. We don't find prevailing colour ofthe different rocks and three cor­ any indication concerning their eventual location and responding facieses. During the initial ground edifica­ period of activity. tion, many quarries were used and most cobblestones had probably not the same aspect. We can estimate Calca ires à entroques (lower Bajocian): the total red cobblestones volume about 3200 cubic This sedimentary bed named, Calca ires à entroques, 2 meters (92.5% from 11300 m , 30 cm thickness). This is irnrnediately placed above the previous (Calcaires volume is toa high for a constant production in those sab/eux) . There is no real evident limit between those geological beds. two calcareous layers and their aspects can sometimes The lower Bajocian doesn't allow any good rock be similar. Nevertheless, the frequently encountered for quality building stones. As we can see in the next rock is a massive encrinite as a result of an accumula­ section, the mechanical properties ofthose limestones tion between marine animal fragments (ossicles) and are not sufficient for actual technical applications. cemented by calcium carbonates (calcite). In the past, two ancient important quarries existed in this geological formation and are now localities, 5 SUBSTITUTION MATERIALS FOR THE Vandoeuvre, and Dommartemont. The first one is RESTORATION located on a cadastral survey from 1805 (50 years after the ground edification) while the second was in The use ofthe original black and red limestones raises use during the 14th century for the construction of a a lot of issues for the remodelling of the Stanislas gate on the Meurthe. Our field observations especially Square. We shall remember that the cobblestones were

269 carefully chosen before their use and despite those pre­ L is a measure of lightness of an object, and ranges cautions, the original pavement was rapidly damaged from O(black) to 100 (white). It indicates the reflecting (see Section 2.4). properties of an object and also reveals its more or less In first instance difficulties appear into making the dark surface. different cobblestones. It would be very difficult to a is a measure of redness (positive a) or greenness dress cobblestones in the ochre brown limestone from (negative a). the lower Bajocian, at reasonable cost. Furthermore, bis a measure of yellowness (positive b) or blueness the existing quarries open in those limestones pro­ (nega tive b). duce crushed granulates and consequently use a very C is a measure of chroma (saturation), and repre­ different quarrying technique that the one used for sents the distance from the neutral axis. the cutting stones. Furthermore, greater difficulties h is a measure ofhue and is represented as an angle 0 0 stand with regards to the durability of the cobble­ ranging from 0° to 360 • Angles that range from 0 stones in relation to lhe weather or occasionally heavy to 90° are reds, oranges and yellows. 900 to 180° are weight vehicles circulation (special events, march yellows, yellow-greens and greens. 180 0 to 2700 are past). Most ofthose limestones are easily cracked by greens, cyans (blue-greens) and blues. From 270° to lhe frost and show bad impact and attrition strength, 3600 are blues, purples, magentas, and return again except the encrinitic limestone, Calca ires à entroques, to reds. for acceptable characteristics. These limestones full of iron oxides would tend to change and generally 6.2 The samples and results choice become red with high temperature. At times, politi­ cal demonstrations takes place on the Stanislas square Eight representative samples (noted A to B) were and it is necessary to anti cipa te, an eventual fire started chosen in the different geological sedimentary beds by the demonstrators and its consequences on the having supplied the cobblestones. Most ofthose sam­ ground. pies carne from the reddish rocks (Aalenian and lower For those above reasons, the local authorities and Bajocian) with reference to the problem ofthe red cob­ the Chief Architect aI Historical Monuments decide to blestones. The black limestone has already the same choose stronger materiais such as better quality lime­ aspect (colour and structure), so only one measure was stones or siliceous rocks. These substitution materiais necessary. will have to be in conformity with the norms and will For each sample, a polished plane surface was have to match to the colour of the original pavement. made and several tests were carried on it. The above The previous descriplions gave any qualitative colours Tables 2a, b contains the obtained values. descriptions that greatly depend on personal percep­ tion. To avoid this mistake, we carried some specific 6.3 Reading analyses: colour of representative samples had been measured with an accurate chromatic study. At first, it is not surprising (but normal) to see a sig­ nificant difference between the results of the black limestone (sample H) and those ofthe red ones. Param­ 6 COLORlMETRlC ANALYSES eters a, b, C and h respective values are -0.29, 0.45 , 0.53 and 123.30, to 8.19,17.90,19.73 and65.54forthe 6.1 Methodology average values ofthe red rocks. On the other hanel, it's difficult to clearly characterize the differences between The colorimetric analyses were carried with two types the red rock sample values. of three dimensional colour spaces, Lab and LCh, That is why we express I1Eab , the difference or the defined by the CIE, International Commission on distance between two colours, in the CIE Lab system, Il/umination (CIE 1986). with the above equation 1: CIE Lab system is the most commonly used colour space and is based on human perception of colou r; the three colour receptors being reei, green and (I) blue in the eye. This result in three sets of signals are sent to the brain: light or dark, red or green, and yellow where I1Eab is the colours difference; I1L the differ­ or blue. They are opposite meaning one receives a red ence between the L values ofthe two different colours; signal or a green one but not both. This opponent type idem for l1a and I1b, with a and b values. colour space is derived mathematically from the CIE From O to 1, the colours difference is tiny. If this values. value is above 3, the visual difference is not very CIE LCh system is similar lo the CIE Lab colour flaggering. Beyonel, the difference becomes more and space, except that it describes the location of a colour more visible. in space by using polar coordinates, rather than rect­ Ali the values between each pair of samples have angular coordinates. been calculated and figured in the next table (Tab. 3).

270 Table 2a. Lab, Colorimetric results. Table 4. Average values for the two cobblestones colours.

Colorimetric analyses Average colorimetric values

Sedimentary beds Samples L a b Colour L a b C h

Calca ires à entroques A 52.43 7.49 17.69 Red limestone* 53.42 7.15 17.28 18.69 67.53 B 56.92 6.60 16.99 Black limestone 41.31 -0.29 0.45 0.53 123.30 C 51.53 6.22 14.52 * limited lO lhe Calca ires sableux and Ca/caires à en/roqlles Calcaire sableux 52.62 16.91 D 6.90 sedimentary beds. E 53.59 8.52 20.28 Minette F 45.72 10.04 21.85 G 42.67 11.58 17.06 as a reference. Indeed we regard (see Section 4.1) this geological leveI as a homogeneous one. Calca ires à G/yphées H 41.31 - 0.29 0.45 Now we can compare the different colours differ­ ences 6.Eab average values calculated between each formation. For the two sedimentary beds Ca/caires à Table 2b. LCh, Colorimetric results. entroques and Ca/caires sab/eux, the average differ­ ence is only 3.77. This value is approximately on the Colorimetric analyses same scale of size that the ones previously obtained Sedimentary beds Samples L C h between the samples of an individual formation. We can then rely on the colours difference in and between Ca/atires à en/roques A 52.43 19.21 67.06 those two sedimentary beds as equivalent. It is not B 56.92 18.22 68.76 the case with the Minette Lorraine, where the typical C 51.53 15.79 66.82 colours differences 6.Eab vary between 8 and 15. Calca ire sab/ellx D 52.62 18.26 67.81 Summarize those colour analyses, we can note that E 53 .59 21.99 67.21 the two sedimentary beds named Ca/caires à entro­ Minelle F 45.72 24.05 65.32 ques and Ca/caires sab/eux are relatively similar. Their G 42.67 20.61 55.83 average parameters (Tab. 4) could be used as a colori­ Calca ires à Gryphées H 41.31 0.53 123.3 metric reference for the choice of alternative materiaIs similar to the original colours of the Stanislas Square pavement. This colour represent a good arrangement as it takes in account most ofthe lithofacies that could Table 3. Colour di fferences !'J.Eab between the different be initially used on the eighteenth century. Whatever samples of red rock. taken decision as whether to use a regular or a spotty ground aspect, the general new aspect will be relatively A B C D E F G close to the historical colour. A O 4.63 3.63 1.00 3.02 8.30 10.60 B O 5.94 4.30 5.05 12.70 15.09 7 CONCLUSION C O 2.71 6.54 10.10 10.66 D O 3.86 9.04 10.99 Upon their accuracy the newly discovered archives E O 8.16 11.78 F O 5.88 completed with some geological investigations allow G O a real knowledge enrichment for the project and constitution specifications. This study allows to find the historical red and At first, we can comment the results obtained black colours, the geometric design ofthe pavement as between the samples taken in the same sedimentary well as information about the growth and cobblestones bed. For the Ca/caires à enfroques (three samples), the hewing or installation. The original black limestones values vary between 3.53 and 5.94. In the formations used for cobblestones are precisely defined while a Ca/caires sab/eux and Minette Lorraine (two sam­ small doubt remains between two adjacent sedimen­ plings, each) the respective colour difference values tary beds for the red cobblestones. The red colour are 3.86 and 5.88. The differences 6.Eab are superior to that is in fact an ochre brown tone, could be from a 3 and some colour variations appear as it was described homogeneous aspect or a fine spotty one. However, in the petrography (see Section 4.2). The smallest a careful observation of the colours linked with an value, 3.86, corresponds to the Ca/caires sab/eux, the accurate chromatic study (CIE Lab and LCh methods, most homogeneous sedimentary bed. reflexion spectrometry) indicate that ali the different Regarding the values of the black limestones, reddish rocks likely to have been used are relatively Ca/caires à g/yphées, it is possible to take those results close in term of colour.

271 This research also gives a clear indication of the Archives 1789. Série AA nO15. Public Record Office poor mechanical properties of the initial materiais. ofNancy. Consequently, the shortly true historical restoration Bubenicek L. 1961. Recherches sur la constitution et la will have to use stones of greater quality that allow repartition du minerai de fer dans I' Aalénien de Lorraine. Annales de ,'ENSG et du CRPG tome 8: pp 5-204. the elaboration of a pavement globally similar to the eastaing J. 1972. Contribution à l'étude sédiment%gique original one. du Bajocien de la région de Nancy. Ph-D Université Henri The conclusion of this study is useful and can be Poincaré, Nancy I. used as a basis for a new substitution materiais choice. eIE 1986. C%rimetry, Technical report /5.2. Vienna: Bureau central de la elE. Pfister C. 1974. Histoire de Nancy. Paris: Berger-Levrault Ed. REFERENCES Zennir A. 1995. Bétons ca/caires en Lorraine - Utilisation des granulats du Bajocien de pour la formula­ Archives 1755. Série CC 630- 691. Public Record Office tion de bétons courants. Ph-D Université Henri Poincaré, ofNancy. Nancy I.

272