Scholarly article on topic 'Historical plasterwork techniques inspire new formulations'

Historical plasterwork techniques inspire new formulations Academic research paper on "Earth and related environmental sciences"

CC BY
0
0
Share paper
Keywords
{Stuccowork / Plasterwork / Density / "Mechanical tests" / "Water absorption"}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — Eunice Salavessa, Said Jalali, Luís M.O. Sousa, Lisete Fernandes, Ana Maria Duarte

Abstract This study is inspired by a revival of traditional stuccoes, plasterwork and recent research on plastering. It includes descriptions of old and new material production techniques with an expected long service life to minimize waste production. A detailed review of the stuccoes and plasters from historical construction treatises was carried out. Their constituents, as well as their functions were studied, and an equivalent eco-mortar was formulated. XRD, EDS, MIP, BET and XRF analysis of the mortars components provided their mineralogical, microstructure and chemical characterization. Petrographic analysis supplied data on the mortars voids. The effect of waste marble and limestone dust on physic-mechanical properties of the eco-mortar was studied in order to find better solutions for restoration interventions. The elastic modulus and compressive and flexural strength of each plaster layer was correlated with the cracking prevention capacity of the mortar. The coefficient of water absorption was estimated in order to study the water exchange behavior between the last two layers of stuccoes and plasters. Furthermore, their behavior with respect to humidity was studied.

Academic research paper on topic "Historical plasterwork techniques inspire new formulations"

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Historical plasterwork techniques inspire new formulations

CrossMark

Eunice Salavessa3'*, Said JalaliLuis M.O. Sousac, Lisete Fernandes, Ana Maria Duarte

a University of Tras-os-Montes e Alto Douro, CITAB/Engineering Department, 5001-801 Vila Real, Portugal b University ofMinho, C-TAC-UM/Department of Civil Engineering, Campus de Azurem, 4810-006 Guimaräes, Portugal c University of Tras-os-Montes e Alto Douro, CGEO-UC/Geology Department, 5001-801 Vila Real, Portugal d University of Tras-os-Montes e Alto Douro, UME/Chemistry Department, 5001-801 Vila Real, Portugal eLNEC/LERevPa, Laboratorio Nacional de Engenharia Civil, Av. do Brasil, No. 101, 1700-066 Lisboa, Portugal

HIGHLIGHTS

• Lime putty, gypsum, rabbit skin glue, sand, limestone fines and marble dust were used.

• 100% Substitution of sand by limestone fines of an "Escaiola" is studied for the first time.

• Marble cut waste in mortar increases the strength and durability.

ARTICLE INFO

ABSTRACT

Article history:

Received 16 January 2013

Received in revised form 29 June 2013

Accepted 21 July 2013

Available online 24 August 2013

Keywords: Stuccowork Plasterwork Density

Mechanical tests Water absorption

This study is inspired by a revival of traditional stuccoes, plasterwork and recent research on plastering. It includes descriptions of old and new material production techniques with an expected long service life to minimize waste production. A detailed review of the stuccoes and plasters from historical construction treatises was carried out. Their constituents, as well as their functions were studied, and an equivalent eco-mortar was formulated. XRD, EDS, MIP, BET and XRF analysis of the mortars components provided their mineralogical, microstructure and chemical characterization. Petrographic analysis supplied data on the mortars voids. The effect of waste marble and limestone dust on physic-mechanical properties of the eco-mortar was studied in order to find better solutions for restoration interventions. The elastic modulus and compressive and flexural strength of each plaster layer was correlated with the cracking prevention capacity of the mortar. The coefficient of water absorption was estimated in order to study the water exchange behavior between the last two layers of stuccoes and plasters. Furthermore, their behavior with respect to humidity was studied.

© 2013 The Authors. Published by Elsevier Ltd. All rights reserved.

1. Introduction

The aim of this study is to clarify the interpretation methods used for ancient plaster formulations, based on a historical overview of construction treatises written throughout the centuries by different authors, up to when it became scientific. The knowledge of traditional and contemporary industrial production is important in both restoration and repair interventions because it enables us to determine the intentions of old designers and master builders, to fit and approach the contemporary mortars to those used in historical masonries and lathing, and to evaluate historic mortars and their formulations in laboratory for assessing their performance and mix optimization.

q This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

* Corresponding author. Tel.: +351 259 350 398; fax: +351 259 350 356. E-mail address: eunicesalavessa@sapo.pt (E. Salavessa).

Lime and gypsum have been used as binders in historical mortars since ancient times. Many ingredients have been added in different periods and regions for fattening the stucco, retarding the setting, and regulating shrinkage and cracking, such as rye flour, rice gluten, burnt gypsum, pork lard, curdled milk, fig juice, albumen, malt and other saccharine or glutinous matters.

Certain industrial wastes and unprocessed micro-fillers, such as marble dust, have been found to provide comparable compressive strength to the control mortars or even to improve it [1]. The use of burnt stone slurry in mortars improves mechanical properties and decreases porosity of concrete mixtures [2]. Pigments present a strong resistance to atmospheric corrosion [3] and improve the compressive strength of blended mortars [4].

Ancient builders generally used a multi-layer system of renders with the external layer characterized by the presence of a significant number of small pores and the internal layer by few, but larger pores. This increased the resistance of the external layer to rain water penetration and prevented the accumulation of moisture inside the masonry or other types of support. In these renders and plasters, the layers not only allowed for water absorption, but

0950-0618/$ - see front matter © 2013 The Authors. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2013.07.064

also facilitated quick evaporation. The render layers in contact with water drainage devices included pozzolanic material with good water resistance [5].

Gypsum mortars composed of gypsum, slaked lime and gravel were frequently used in Medieval Spanish architecture. In addition to making the mortar harder and more resistant to damp, the car-bonation of lime also retards the mortar setting to allow more time to decorate the reliefs. Gravel makes the mortar more porous, while glue and salt retard its setting. Plasterwork was usually composed of three layers: first, a thick layer of mortar where the decorative motives were outlined; next, the second mortar layer was carved; finally, the last layer was a whitewash base to cover the plastered surface [6]. Towards the end of the Middle Ages and the beginning of the Renaissance, the technique of the "fresco" was described by Cennini, who recommends that the mortar must be composed of 2 volumes of sand and 1 volume of lime [7]. In case a ''rendering coat'' was needed on the masonry, this was constituted of a mortar with coarser sand to facilitate the adherence of the ''setting coat'' on which the work was painted [8].

During the period of Raphael and the great revival of stucco-work, a practical recipe for stucco was tried by Master Jacopo de Monte St. Saviano the Sculptor, written around 1503. The stucco for making and modeling figures, coloring them and providing water resistance is formulated with 2.5 volumes of fine pounded marble and 1 volume of slaked lime which must be stirred and beaten together into a fine paste, then either formed by hand or in a mould and cured in the shade [9].

A mixture described by Vasari included 2:1 (in volume) of lime and marble dust, to which a little gypsum was added to set the plaster. Pirro Ligorio, an architect and coadjutor of Michelangelo for St. Peter's at Rome, wrote a recipe using 3 volumes of pounded Parian marble from the ruins in Rome and from broken statues, and 1 volume of perfectly slaked lime [9]. Palladio recommends a mortar with a proportion of 1:3, i.e., 1volume of lime to 3 volumes of mine sand. If the sand is from a river or the sea, the proportion must be 1:2 [10]. The pozzolana from volcanoes north of Rome, which have thin glassy fragments as an active component, are found in the fresco mortar of''The Last Judgment'' by Michelangelo in the Sistine Chapel [11].

The slaking operation is a rapid exothermic reaction of calcium oxide with water which releases heat and produces calcium hydroxide. Calcium hydroxide then reacts with carbon dioxide, forming calcium carbonate [12]. The rate of carbonation is high in an atmosphere with low concentrations of CO2, low temperatures and a high relative humidity [13]. In both Nuremberg and Zurich, burnt lime was stored dry and sold as quicklime, which was then slaked before its use in the construction process during the 16th and 17th centuries. Pieces of quicklime were reduced to slaked lime powder, when sufficient water was added to change the CaO into Ca(OH)2. When excess water is added, the quicklime becomes lime putty [14].

Plasters as described in 17th and 18th century treatises are characterized by their technological complexity, due to the composition of several layers with different microstructures and their role in regulating the movement of humidity towards the exterior layers [15]. Italian baroque stuccoes were made with 3 layers: first the bulk, with its hydraulic properties; the 2nd layer, which was more plastic for decorations made of calcic and dolomitic lime and a sand mortar with good mechanical strength, or a gypsum mortar without aggregates; lastly, the finishing mortar with a few layers of gypsum [16].

Rondelet, referring to 18th century stucco, indicates that the mixture for ornaments must have one part of lime to one part of marble dust. After a mortar of lime putty with marble dust was prepared, the ''floating coat'' was wetted and the stucco was applied with a brush, spread with a spatula, and finally polished with

a steel mason's chisel and a damp linen cloth. For the final layers of plain surfaces, two parts of lime and one part of marble dust was necessary for stucco. For exterior facades, or in wet environment, pozzolana or tile dust was added to the ''floating coat'' [17].

In the 19th century, the 7-10 mm thick ''bulk'' layer usually had one of the following mixtures: lime/gypsum binder and sand; a coarse and rugous mortar of lime/gypsum and pozzolana; or, as Leitao [18] indicated, a lime/coarse sand mortar (1.5:1). The next 1-to-2-mm-thick layer was a mixture rich in gypsum or lime, using a low percentage of fine aggregates for moulding and decorative elements, and the exterior surface was made entirely of hydraulic lime mortar. The final 1-to-3-mm-thick finishing layer was made of gypsum paste, lime/gypsum paste, or lime/gypsum paste and marble dust [19]. Luis Leitao states that late 19th century plasterers employed two mortars, the ''floating coat'' and the ''setting coat''; the first one was laid directly on the ''bulk'', composed of one part of lime putty, one part of gypsum dust and four parts of limestone sand; the second layer was spread on the ''floating coat'' and consisted of equal parts of lime putty and gypsum [18].

Pigments were added to the setting coat in order to imitate the marble, or Italian ''Scagliola'', and to achieve the ''base'' color. The method to produce marble veins consisted of thin cakes of a mixture of gypsum, pigments and glue water, which were cut into strips and then expanded with skimming float over the colored setting coat [17]. According to Luis Leitao, the ''Escaiola'' is a mortar made of lime putty, white stone dust and fine washed sand, in equal parts, to which pigments may be added. For the exterior facades, the stone dust was substituted by cement [18].

Plaster with fibers was patented in 1856 by a French modeler, Leonard Desachy. It was mainly used for casting purposes and corresponds to manufacturing slabs, casings and other forms by combining calcium sulfate, burnt or boiled gypsum, or plaster of Paris with jute woven into an open mesh and strengthened or strutted with wood [20].

The objectives of this study are: (1) to evaluate lime-or-gyp-sum-based mortars potentialities which incorporate waste marble powder and limestone fines; and (2) to study the advantages of the mixed-gypsum mortars in the decorative plasterwork. The evaluation of lime-putty-based mortars with a high volume of fines is of special interest due to their similarities to historical ones. Furthermore, such mixes are durable system renders and mortars prescribed in historical construction treatises also present better mechanical, physical and chemical compatibility with the old masonries and lathing they cover. This is essential aspect when undertaking repair interventions.

In this study it was considered important to reproduce some stuccoes and plasters formulated in ancient treatises as a study tool, in spite of lacking the exact correspondence. These mortars must have lower mechanical properties than the support where they are applied. Interior plasters and stuccoes must be permeable to moisture, thus improving water migration towards the exterior. The performance must be similar to that of ancient mortars. Their principle working properties are important as they are needed in order to evaluate their performance. When ancient texts are translated, some technological questions remain to be resolved in regard to the technical terminology or technological and scientific content. To verify if translations of ancient texts are correct, an experimental reconstruction is needed in order to determine whether the indications given by the authors were true to old construction traditions. Therefore, mortars as similar as possible to historical mortars that have shown desirable longevity and aesthetic aspects were produced for this purpose. This is a necessary step for mortar formulation in renovation and repair interventions. It also serves for the formulation of new mortars with similar characteristics.

In this research, the effect of the elastic modulus, the compressive and flexural strength of each plaster layer on crack prevention, including the layer under the decorative lathwork plaster of the ceiling is evaluated. Furthermore, the coefficient of water absorption estimated from a capillary suction test provides the necessary information for water exchange capacity between the second and third layers. This way, the humidity exchange behavior of stuccoes and plasters can be evaluated.

2. Materials and methods

2.1. Mortar components and composition

In order to evaluate the interior stuccoes and plasters, the following 7 mortars were produced based on the materials and technologies referred to in the recorded treatises. Table 1 indicates the type of mortars according to their binders and applications; Table 2 indicates the bulk density, chemical composition, specific surface area and porosity of the components used; Table 3 shows the particles size distribution of the aggregates; Table 4 indicates the mortars studied.

In addition, a general description of the 7 mortars studied in this study is provided here.

(1) ''Cennini's Intonaco" (lime and sand), similar to the "intonaco" indicated by Cennini (15th century); this mix was used as the control mortar to be compared with ''Leitao's Escaiola'' and ''Stucco II" to evaluate the potential utilization of waste marble sawdust and limestone dust, and how these industrial by-products influence the eco-mortars performance in their physic-mechanical properties;

(2) ''Rondelet's Stucco III" (lime and marble dust), similar to the mortar indicated by Rondelet (18th century), with a larger proportion of lime to improve strength and fresh-state workability for ornamental mouldings;

(3) ''Leitao's Escaiola'' (lime, marble powder and sand), similar to the ''Escaiola'' indicated by Luis Leitao (1896), but with less calcite than the latter due to its application in plain plasters with no particular workability requirement. Due to ecological considerations, dark yellow cadmium was added to produce a decorative stucco with a base color. The color is similar to the ''almagre'' found in Leitao's color catalogue, which is a clay produced in St8. Maria, in the Azores [21]. Yellow cadmium was discovered in 1818 by Stromayer and the red cadmium appears in 1926 [22] and is sold nowadays on the market as ''almagre'';

(4) ''Stucco II/Proposal'' is an eco-mortar proposed in this research work, using lime putty, marble powder, limestone dust and cadmium. This mix is formulated to be compared with Leitao's Escaiola, replacing the river sand by a limestone filler;

(5) ''Leitao's Stucco I/Floating Coat'' (lime, gypsum, limestone fines), for the 2nd layer application which covers the rendering or bulk layer;

(6) ''Leitao's Setting Coat'' (lime and gypsum); for the finishing coat, in contact with the exterior atmosphere.

(7) ''Leitao's Scagliola'' (gypsum, glue water and blue ultramarine pigment); for the finishing coat, in contact with the atmosphere.

Materials used in this study were from Portuguese sources, as indicated below:

(1) Lime putty is a commercial natural air lime paste of Palhais, Charneca da Caparica, class CL90 according to EN 459-1 (D. Fradique) with 1.5 cm car-bonation in 240 days; this lime putty is composed of calcium hydroxide,

calcium carbonate, magnesium hydroxide and magnesium oxide and contains 48% solids by mass; DRX analysis revealed portlandite and calcite; the median pore diameter is 0.2177 im;

(2) Two types of commercial natural gypsum of Leiria, Souto da Carpalhosa (CaSO42H2O; Sival) were burned to prepare the setting gypsum; the water/binder ratio used was 0.8; the coarse gypsum which DRX analysis reveals bassanite, calcite and calcic sulfate has a light grey color due to the presence of clay, and the fine gypsum (bassanite and calcite) is a white-colored dust; the median pore diameters of coarse and fine gypsum are 0.3823 im and 0.2355 im, respectively;

(3) Rabbit glue, a proteinaceous and adhesive substance, obtained by boiling young rabbit bones, skin and cartilages in water in closed vessels, is employed as a bonding additive to the water for making the "Scagliola"; the glue assists in the smoothing process of plaster surfaces;

(4) White marble powder from a sub-crystalline calcareous rock of Pero Pinhe-iro; DRX analysis revealed quartz and calcite; the pore diameter of marble dust is 0.2520 im;

(5) Grey to brown limestone fines in which DRX analysis detected calcite, feldspar and quartz, from a marly limestone quarry at Sesimbra, with intrusion of volcanic matter, (aggregate; particles size Gr 85 according EN 12620: 2000 +A1: 2008; José Marques Gomes Galo); part of it was transformed in filler; the darkest particles of the grey to brown fines are volcanic particles; the median pore diameter is 0.1976 im;

(6) Fine white to grey siliceous river sand from Bucelas; the median pore diameter is 0.9442 im;

(7) Ultramarine blue of Kremer, an inorganic pigment manufactured by burning sodium sulfate, quartz, and sulfur around 800 "C, which produces a sodium aluminium silicate sulfide structure (Na8Al6Si6O24S2); the DRX analysis revealed the presence of kaolinite, ilite, quartz and ultramarine blue pigment; the ultramarine blue median pore diameter is 0.2766 im;

(8) Cadmium "sulphoselenide" is a mixture of cadmium sulfide (CdS) and selenium (Se) and barium sulfate (Ba SO4) with a dark yellow color, the shade of which is due to the amount of selenium; the DRX analysis revealed calcite, hematite and the cadmium pigment; the median pore diameter is 0.3175 im.

For simplicity and accuracy in preparing mixes, the volume proportions of components indicated in the old treatises were converted to weight proportions. In total, three binding systems were used: non-hydraulic lime putty; non-hydraulic lime putty-plaster; and plaster of Paris-rabbit glue water. The aggregates used were natural quartz sand, sand with 41.7% limestone fines; limestone dust and marble powder.

2.2. Mortar preparation

In the studied lime based mortars, the water/binder (H2O/Ca(OH)2) mass ratio (W/B) was calculated from the dry mass Ca(OH)2 of the lime putty, which was measured to be 480 kg of quicklime for 1 m3 of lime putty. This resulted in a W/B ratio equal to 1.08 for the calcic lime putty (fat lime), according to EN 459-1 [23]. Mortar mixtures were prepared with the required amount of water to obtain a normal consistency and a suitable workability using manual or mechanical procedures. The workability was measured by a flow table according to EN1015-3:1999 [24]. The workability values for ''Stucco II/Proposal'' and ''Scagliola'' are presented in Table 5 [18,25,26]. Mortars specimens were prepared and compacted in 40 x 40 x 160 mm prismatic stainless steel moulds with a glass rod to remove air bubbles. Specimens were cured at a temperature of 20 ± 2 "C and 65 ± 5% relative humidity for 90 days. All the specimens were sufficiently hard to be removed from

Table 1

Composition of the studied mortars according to their type/binder and structural function.

Mortar W/B Type of mortar Type of binder Aggregate Application

Cennini's Intonaco W/B = 0.82 Non-hydraulic Non-hydraulic lime Sand 2nd layer of frescoes (to be covered by Intonachino with

lime mortar putty better mechanical performance)

Rondelet's Stucco III W/B = 1.03 Non-hydraulic Non-hydraulic lime Marble powder Ornamental mouldings of interior decorative stuccoes; better

lime mortar putty aesthetic effect

Leitao's Escaiola W/B = 1.28 Non-hydraulic Non-hydraulic lime Marble powder; sand; Final layer of interior decorative stuccoes; better aesthetic

lime mortar putty cadmium effect

Stucco II (Proposal) W/B = 1,28 Non-hydraulic Non-hydraulic lime Marble powder; 2nd Layer of interior decorative stuccoes; final layer of

lime mortar putty limestone filler; interior decorative stuccoes

cadmium

Stucco I (Float. Coat) W/B = 0.80 Gypsum-lime Lime putty and Limestone fines 2nd Layer of interior decorative plasters

mortar anhydrous gypsum

Leitao's Setting Coat W/B = 0.81 Gypsum-lime Lime putty and Final layer of interior decorative plasters

mortar plaster of Paris

Leitao's Scagliola W/B = 0.74 Gypsum mortar Plaster of Paris; Ultramarine blue Final layer of interior decorative plasters; better aesthetic

rabbit glue effect

Table 2

Bulk density, chemical composition (mass%), BET and porosity of components.

Lime Brown. Plaster of Rabbit Sand Marble Limestone Ultramarine Yellow of

putty plaster Paris glue fines powder fines blue cadmium

Density (kg/cm3) 1255 670 698 623 1422 771 1518 (sand) 2250 4820

1167 (filler)

CaCO3 a 99.2 99.0 85.7

CaSO4 96.9 93.3

SO3 5.6 0.7 23.1

CaO 4.4 27.2 0.2

MgO 0.2 1.1 0.5 0.5 0.4 1.4 0.2

SiO2 0.3 1.3 1.0 0.2 94.7 0.3 5.6 32.5

Al2O3 0.1 0.5 0.4 0.1 4.2 0.1 2.4 22.8

Fe2O3 0.7 3.0 0.2

Na2O 4.6 19.5

K2O 0.2 1.2

TiO2 0.9 0.5

Cl2O 60.2

P2O5 0.3

Sr 0.1

Si 0.1

S 0.9

Fe 8.6

Cd 5.1

Ca 52.5

Ba 2.6

Se 1.3

Pb 0.2

Pd 0.2

SBET (m2/g)b 14.24 4.397 3.916 6.335 0.3825 1.206 2.721 4.068 1.353

Porosity (%) 61.15 59.81 38.09 15.95 40.82 51.50 45.84 61.42 49.05

a Gravimetric analysis. b SBET (BET specific surface area).

the moulds after 48 h, with the exception of ''Scagliola'' that was removed after 72 h. This was due to the fact that both the glue water and the ultramarine blue used in this mix are setting retarders. The specimens were kept in a controlled environment until the time of testing in appropriate conditions to facilitate the carbonation of calcium hydroxide.

A minimum of six specimens were tested after a 90-days carbonation period for each of the seven mortars studied. The dimensional and weight variations, the dynamic modulus of elasticity, flexural and compressive strength, as well as, capillary water absorption were evaluated according to EN 1015-11:1999 [27], EN 132792:2004 [28] and EN 1015-18:2002 [29].

2.3. Analytical techniques, tests performed and equipment used

Mineralogical, morphological, microstructure and chemical analysis of the different component materials were made. Samples were submitted for X-ray diffraction (XRD), by PAN'alytical, X'Pert Pro powder diffractometer with X'Celerator detector, to study the mineral components. Samples were prepared in a standard samples bearer using a spinner stage. 40 kV and 30 mA were used and the procedure was done in Bragg-Bentano geometry, between 7° < 2_ < 80. Energy dispersive spectroscopy (EDS) allowed for the chemical analysis of the elements present in the samples and, in conjunction with the images obtained by scanning electron microscopy (SEM, by Philips-FEI/Quanta400 with EDS), it was possible to carry out the characterization of the crystalline structure of the materials. Mercury intrusion porosimetry (MIP) provides total porosity (%) and pore diameter average (im), by Auto Pore IV of Micrometrics (0.5 PSI - 33.000 PSI). Specific surface area (m2/g) was determined by Brunauer-Emmet-Teller (BET) method, with physical adsorpsion isotherm data. The grain size distribution of aggregates was obtained by sieving, according to EN 1015-1 and EN 13279-2. The filler size was estimated using an Alpine vibrator.

The squeeze flow testing was done according to EN 1015-2, EN 1015-3 and EN 13279-2. The metallic plate was 300 mm in diameter; the mould ring 60 mm high, 100 mm in diameter at the base and 70 mm at the top. The flow-table was shaken 15 times, with one stroke per second. The apparent bulk density of the powder or putty materials was determined by the quotient between the mass and the volume after compaction, according to NP EN 1097-6, EN 1015-6 and EN 1015-2. Average values of weight and dimension variation of the specimens were registered 90 days after casting and were used to evaluate the water loss during this period [30].

It was observed that carbonation process leads to an increase in weight of non-hydraulic lime binder of 35%. The carbonation of the lime putty represents the transformation of calcium hydroxide into calcium carbonate in the presence of

Table 3

Particle size distribution of the aggregates.

Particle size White sand Marble Limestone Limestone

(mm) fines powder fines filler

Passing (%) Passing (%) Passing (%) Passing (%)

6.3 - - 100 -

4.0 - - 95 -

2.0 - - 63 -

1.0 - - 17 -

0.500 - - 24 -

0.425 99.5 - - -

0.300 55.0 - - -

0.250 - - 15 -

0.212 60.0 - - 90.38

0.180 - - - 87.48

0.150 89.0 - - 81.86

<0.150 96.5 - - -

0.125 - - 11 79.88

0.090 - - - 71.7

0.063 - 100 8.3 61.5

<0.063 - - 41.7 -

0.040 - 86.2 - 55.14

0.032 - 89.2 - 47.7

water and carbon dioxide. The carbonation process of 74 g of Ca(OH)2 leads to 100 g of CaCO3 and the evaporation of 18 g of water. This happens because lime captures 44 g of CO2:

74 g Ca(OH)2 + 44 g CO2 = 118 g CaCO3; 118 g CaCO3 - 18 g H2O = 100 g CaCO3

The dynamic modulus test was performed in order to evaluate the cracking susceptibility of different stucco or plaster layer of a given thickness covering masonry or lather support. The lower modulus of elasticity of a plaster indicates a lower probability of cracking [31-33].

Three samples of each mortar were submitted to flexural and compressive strength tests, using a 50 N/s and 100 N/s rate of loading respectively, and the average values are estimated and reported here.

Table 4

Mortar mixes studied.

Mortar/author Components Weight (g) W/B ratio B/aggreg. ratio

Cennini's Intonaco3 Lime putty 1046 1.08 1 Lime putty

White sand fines 2370 2 River sand

Rondelet's Stucco III Lime putty 2092 1.33 2 Lime putty

Marble powder 643 1 Marble powder

Leitao's Escaiola Lime putty 1046 1.58 1 Lime putty

Marble powder 642 1 Marble powder

White sand fines 1185 1 River sand

Dark yellow cadmium 20

Stucco II/Proposal Lime putty 1046 1.58 1 Lime putty

Marble powder 642 1 Marble powder

Limestone filler 1265 1 Limestone filler

Dark yellow cadmium 20

Leitao's Stucco I (Floating Coat) Lime putty 448 0.93 1 Lime putty

Browning plaster 239 1 Browning plaster

Limestone fines 2169 4 Limestone fines <0.032 mm

Leitao's Setting Coat Lime putty 1568 0.93 1 Lime putty

Plaster of Paris 873 1 Plaster of Paris

Water 698

Leitao's Scagliola Plaster of Paris 873 0.74 1 Plaster of Paris

Rabbit glue 69 0.88 Glue water

Ultramarine blue 20 0.02 Pigments

Note: W/B ratio (or W/Ca(OH)2 ratio has been calculated from mass of dry Ca(OH)2 of lime putty which contains 48% solids by mass. a This mix showed low workability.

Table 5

Workability of Stucco II and Scagliola, measured by flow table (EN 1015-3:1999).

Mixture

W/B Workability Observations

Stucco II/ Proposal

Scagliola

1.08 185 mm

0.74 165 ±5 mm

The W/B ratio and workability were measured by flow table to achieve reference value of 175 ± 10 mm. The proposed mortar was inspired by the Escaiola mortar, with 100% substitution of sand by limestone filler

The W/B ratio is higher due to smaller particle size of marble powder [26,27] and limestone filler, which showed similar effects on reducing viscosity in Stucco II

The indicated workability is questionable as it was not possible to obtain a reliable measurement due to high viscosity and complete adherence of the paste to the mould ring. While the Scagliola mix was prepared manually and took about 4 h, the other mixes were prepared with 1 min of manual mixing and 4 min mechanical mixing. The W/B mass ratio for Scagliola followed the Leitao's recipe [16]. The consistency of this mortar is plastic and can be easily spread and moulded for fast and flawless application

The capillary water absorption is measured by the mass of water absorbed for a given time, by a dry specimen, when the surface is in contact with water. The coefficient of water absorption C, was calculated by the following equation:

M2 - M1 1 i kg \

C S Xptffi-u{ m2mjn05J

where Mj and M2 are the masses (kg) at times tj and t2 (min) and S (m2) is the area of the specimen in contact with water.

The thin sections of the stuccoes and plasters were prepared and examined with a petrographic microscope with eye pieces of 4 magnifications using both natural and polarized light. Petrographic analysis provides information on textural features, mineralogical data, shape and dimension of aggregate grains, fractures and pores of formulated mortars.

3. Discussion of the results

Results obtained from the physical and mechanical tests are shown in Table 6. Figs. 1, 2 and 3 present the petrographic analysis.

3.1. Petrographic analysis

(1) Cennini's Intonaco (Fig. 1-1): sub-round to sub-angular quartz grains (Qz), with very homogeneous size, normally inferior to 0.4 mm are observed. Some carbonated grains (CaCO3) are also visible; the matrix constituted by lime (L) is small in quantity and presents voids (V) with elongated shape (<0.1 mm) or with circular shape (<0.5 mm).

(2) Rondelet's Stucco III (Figs. 1 and2): presents a very fine texture, where areas with lime (L) predominance are distinguished from areas with marble (M) dust predominance; some carbonated fragments (CaCO3) of larger sizes with some round pores which can reach 0.5 mm in diameter are observed. Furthermore, a crack (F) which seems to correspond to material shrinkage is noted, while areas with predominance of lime and others with marble dust in greater quantity are observed, indicating that these two materials are not entirely mixed.

(3) Leitao's Escaiola (Figs. 1-3): quartz (Qz) sub-round and sub-angular fragments are visible, with maximum dimension about 0.5 mm, distributed in a matrix constituted by a lime and marble dust (L+ M) mixture; the darkest matrix color must be due to cadmium; circular voids are frequent (V) with dimension of 0.1 mm and cracks (F) are very frequent; the irregular cracks have maximum length of 1 mm and maximum width of 0.05 mm.

(4) Stucco II/Proposal (Figs. 2-4): in a fine matrix constituted by lime, marble dust and limestone fines (M + L + M), can be distinguish. Some volcanic fragments (Vf) of small size; limestone fragments (L) are also frequent with maximum dimension of about 0.5 mm. Some calcite aggregates (CaCO3), which can correspond to marble fragments, and some oxide spots, which correspond to fragments of volcanic material, are visible. Circular voids (V) less than 0.4 mm are frequent.

Table 6

Experimental results at 90 days (160 mm x 40 mm x 40 mm prismatic specimens).

Mortar water/binder ratio Bulk density Volume Dynamic E- Flexural strength Compres. Coefficient of water absorb.

(kg/m3) variations (%) modulus (MPa) (N/mm2) strength (N/mm2) (kg/m2 min05) 10-90 min

Cennini's Intonaco W/B = 0.82 1651.5 Shrinkage 97.4 3538.4 0.46 1.18 1.41

Rondelet's Stucco III W/B = 1.03 1280.6 Shrinkage 66.9 7910.8 0.15 2.15 3.13

Leitao's Escaiola W/B = 1.28 1663.3 Shrinkage 87.1 4484.5 0.73 1.77 1.97

Stucco II (Proposal) W/B = 1.28 1537.7 Shrinkage 91 3361.1 0.87 2.2 2.42

Leitao's Stucco I (Float. Coat) W/B = 0.80 1548.6 Shrinkage 95.2 3162.7 0.93 2.1 2.43

Leitao's Setting Coat W/B = 0.81 780.5 Expansion 101.7 1537.0 1.47 2.54 8.18

Leitao's Scagliola W/B = 0.74 754.2 Expansion 109.5 1705.7 1.90 1.3 2.93

Note: the unfilled line indicates the separation between lime-based and gypsum-based mortars.

Fig. 1. Microphotographs of formulated stuccoes: (1) Cennini's Intonaco; (2) Rondelet's Stucco III; (3) Leitäo's Escaiola.

Fig. 2. Microphotograph of formulated eco-stucco: (4) Stucco II/Proposal.

(5) Leitäo's Stucco I (Floating Coat) (Figs. 3-5): presents a similar aspect to the preceding mortar (Stucco II/Proposal), except that it presents more volcanic fragments and oxide particles.

(6) Leitäo's Setting Coat (Figs. 3-6): is a homogeneous material constituted by a lime and gypsum agglomerate (L+G), where some carbonated nodules are detached with dimensions normally inferior to 0.2 mm. The remarkable aspect is the presence of many voids (V), lightly elongated and with

irregular contours, with a common dimension of about 0.1 mm. The round voids are less frequent and can reach larger sizes (0.3 mm).

(7) Leitao's Scagliola (Figs. 3-7): is a homogeneous mixture, where high porosity stands out with some pores of large sizes (0.6 mm), while the majority are of small size (0.1 mm). All the black spots in the photo correspond to voids, which in spite of being hard to distinguish, exist in significant quantity. There are some concentrations of carbonated material (CaCO3) of a larger size (0.2 mm), but the majority are smaller in dimension; the gypsum (G) and the carbonates of small dimensions constitute the matrix; the glue (Rg), of dark green color, is distributed through the matrix with concentrations in some areas.

3.2. Shrinkage and bulk density

Shrinkage and expansion were measured in the three axial dimensions and volumes of 90-day prisms. The average values of six specimens for each mortar are shown in Table 6 and Fig. 4.

Cennini's Intonaco with sand has the lowest shrinkage due to the lowest water to binder ratio (W/B ratio). Shrinkage is especially observed in mortars with marble dust due to a higher water/binder

Fig. 3. Microphotographs of plasters: (5) Leitao's Stucco I (Floating Coat); (6) Leitao's Setting Coat; (7) Leitao's Scagliola.

I ■ Decrease Increase |

s? # <> A0' 0-

" -35 I -33 1

Fig. 4. Decrease or increase of volume in the seven hardened mortars, at 90 days.

ratio [25], as in Rondelet's Stucco III. The high ratio of lime putty/ marble dust results in a mortar with high porosity, deformability and shrinkage as observed in Escaiola, and Stucco II. The Stucco II, which substitutes the100% of fine sand in Escaiola by the limestone filler, has a lower shrinkage value, in spite of both mortars having the same water to binder ratio. The substitution of sand by a limestone filler, with the presence of 12.4% of SiO2 + Al2O3 + Fe2O3 + MgO and 85% of CaCO3 and small quantities of TiO2,with hydraulic properties, promoted the residual hydration water loss and restricted hydration. Hydraulic elements, which do not react with lime, act as an inert filler, reducing shrinkage. In fact, although the bulk density of fresh Stucco II is higher than the fresh Escaiola, hardened Stucco II is lighter than Escaiola, which is due to water loss and, hence, lower shrinkage.

The very small pores (below 2 im in diameter) in the matrix, resulting from the evaporation of any excess water not consumed during the hydration process, contribute to a slower carbonation of lime, leading to a decrease in weight of the binder. In Escaiola, the presence of sand allows for a greater volume of pores whose larger diameter promotes access to CO2 [34] and drying shrinkage of the lime. In the Floating Coat, the coarse gypsum decreased shrinkage.

Expansion is observed in mortars with gypsum (Leitao's Setting Coat and Scagliola). In the Setting Coat, the drying shrinkage which characterizes lime is compensated for by the expansion of the gypsum during the hardening process [35]. In Scagliola, the addition of the ultramarine blue pigment reduced the expansion of the gypsum-glue water-based mortar, in part due to the decrease in water absorption [4].

Regarding bulk density, the higher values are of mortars with standard sand (Cennini's Intonaco and Leitao's Escaiola) in comparison with mortars with marble or limestone fillers (Stucco

Time (min1/2)

Fig. 6. Capillary curves for the seven mortars, at 90 days.

II, Rondelet's Stucco III, Floating Coat). As presented in Table 4, where the the sand of Cennini's Intonaco in the Stucco II is substituted 100% by marble dust and limestone dust, the Stucco II has a lower bulk density. When comparing Escaiola with Stucco II with the same water content, the unit weight of the later decreases. The reason is an increase in air content, related to the substitution of sand by limestone filler. In spite of the substitution of 50% of sand of Cennini's Intonaco by marble sawdust in Escaiola, the bulk density is higher in the later due to its higher water content for lubricating filler particles [25]. In the Escaiola and Stucco II, the adsorption of cadmium onto silica of the sand and stone dust, must have contributed to increasing the bulk density of the hardened mortars [36].

Scagliola and Setting Coat with gypsum show the lowest weight values. This property has an important role in the behavior of plastered lathwork ceilings before cracking of the inner layer because mortars which constitute the last layer cannot impose a heavy load [25].

3.3. Dynamic E-modulus, flexural and compressive strength

To decrease the probability of cracking in plaster layers, their dynamic modulus of elasticity and the shrinkage deformations must be decreased [33]. Many plasters and stuccoes are suspended in wooden lathwork ceilings and present different types of cracks. The thickness and E-modulus of each plaster or stucco layer play an important role in the behavior of the ceiling before cracking

3 2.5 2 1.5

OFlex. Str. □ Comp. Str.

Ф □

1) Cennini's Intonaco, Edyn = 3538 MPa; 2) Rondelet's Stucco III, Edyn = 7911 MPa;

3) Leitao's Escaiola, Edyn = 4485MPa; 4) Stucco II, Edyn = 3361 MPa;

5) Floating Coat, Edyn = 3163 MPa; 6) Setting Coat, Edyn = 1537 MPa; 7) Scagliola, Edyn = 1706 MPa

Fig. 5. Flexural and compressive strength results according to dynamic modulus of elasticity of stuccoes and plasters, at 90 days.

• Floating C.

Setting C.

Scagliola

80 70 60 50 ' 40 30 20 10 0

Time (h)

Fig. 7. Drying curves of the three studied plasters (at constant external conditions of temperature, relative humidity and air velocity). Drying curves show that drying rate is similar to Floating Coat and Setting Coat; Scagliola has a larger drying rate curve, because hardly dries.

of the mortar layer. High plaster stiffness reduces tensile stress in the external and lower layer [31].

Mortars which use gypsum have a low dynamic E-modulus (Floating Coat -E = 3163 MPa, Setting Coat - E = 1537 MPa, and Sca-gliola - E = 1706 MPa). In the aerial lime/gypsum mixture of the Floating Coat gypsum accelerates the mortar setting, and in the Setting Coat and Scagliola gypsum makes the mortar stiffer, reducing the tensile stresses and preventing the decorative surface layer from cracking. Thus, the gypsum mortars are suitable for wooden plaster frameworks such as ceilings or partition walls. Rondelet's Stucco III used in renders, cornices and ornamentations has the maximum Edyn value (E = 7911 MPa) due to a higher proportion of lime putty, while showing the lowest flexural strength (see Fig. 5).

The microstructure of Rondelet's Stucco III, with marble dust as a filler material, indicates that a good bond between the aggregate and the lime putty has occurred, thus developing good cohesive-ness to mortar with a high compressive strength. This cohesiveness conferred by the marble powder is a consequence of its high specific surface area [25,26]. Limestone dust and marble powder increase compressive strength due to void filling. The fineness of the marble dust is higher than the limestone dust, thereby conferring good cohesiveness to Escaiola and Stucco II. It is noted that Escaiola, with a 50% substitution of Cennini's Intonaco sand with marble powder, produces a mortar with a better mechanical performance. Comparing Escaiola with Stucco II, the latter presents higher mechanical properties due to the substitution of 100% of Escaiola sand by limestone dust. The compressive strengths obtained for Floating Coat and Stucco II are similar, due to a similar binder/aggregate volume ratio. However, Floating Coat, with limestone fines and a lower water/binder ratio, has a higher flexural strength than Stucco II with a limestone filler, while their compres-sive strengths are almost the same.

Mortars with gypsum have higher flexural strengths than the previous ones. The mortars with the highest values of compressive strength are lime putty-gypsum and then lime putty-stone dust. In the Floating Coat and the Setting Coat the gypsum mixed in the lime putty increases the speed of hardening and assures a higher adherence of the mortar, compensating for the lower adherence of lime. The gypsum-based mortars are much harder and compact than the lime-based mortars [37].

The Setting Coat shows a higher mechanical strength than the Floating Coat, due to the addition of middle sized aggregates such as limestone fines used in the Floating Coat, which increase the empty pores, promoting a rapid carbonation of the mortar [37] and retarding the complete carbonation in depth, to reach higher mechanical resistance. The lowest compressive strengths are observed in Scagliola gypsum and glue water, and in Cennini's Render with lime and sand (Fig. 5).

The Scagliola is a plastic material, ideal for making decorative elements. In this gypsum/glue water mortar, the addition of animal glue increases the flexural strength of the mortar and works as a hardening retarder, acting as a precipitation inhibitor [37] to allow time for the decorative elements to be done and to obtain a better quality mortar.

At 90 days curing time the mechanical strength of the lime-based mortars are lower than that of gypsum-based mortars due to the carbonation process which can continue for many years, further increasing their strength. Lawrence et al., explain how the car-bonation of air lime mortars affects the pore structure of the mortars, producing pores of 0.1 lm in diameter associated with their carbonation front, and smaller pores not involved in the car-bonation process make the core of the mortars carbonate slower, varying with the type of aggregates mixed in the lime mortars [34].

Rondelet's Stucco III in comparison with Escaiola presents a higher compressive strength, and this can be mainly attributed to the

higher lime content in the mortar. In the Escaiola and Stucco II the adsorption of cadmium onto the silica of the sand, marble dust and limestone dust apparently contribute to increasing the tensile strength of the two mortars [36].

In Scagliola, the ultramarine blue pigment has a high capacity of CaSO42H2O fixation and improves the compressive strength of the gypsum-based plaster with glue water. The glue water decreases the compressive strength but increases the flexural strength. Both the glue water and the ultramarine blue are setting retarders. The ultramarine blue, due to its sulfide nature and its high specific surface area, makes the mixture demand more water, causing it to have a long setting time in comparison with Leitao's Setting Coat [4,30].

3.4. Water absorption due to capillary suction

According to D'Orazio and Stazi, in the three-layer system of the historical plasters and stuccoes, the 1st layer has a greater capillary suction capacity to drain water to the outside layers. The 2nd layer has a lower suction capacity than the first one, thus increasing the water content of the 1st layer. Being resistant to the flux of water to the outside, when the 2nd layer, which is considerably thicker, has accumulated the maximum water content, the 3rd layer then transports water to the outside, due to its greater suction capacity and better water accumulation capacity than the 2nd layer [15].

In this study, the highest coefficients of water absorption are found in mortars with gypsum and marble powder (Fig. 6). Compositions like Floating Coat, containing fines such as limestone and marble powder in a greater proportion than lime putty, have a lower rate of water absorption than Setting Coat and Rondelet's Stucco III, because the fines block capillary pores. So Floating Coat and Stucco II are suitable for the 2nd layer of the plaster or stucco covering system, and Setting Coat and Rondelet's Stucco III for the 3rd and last layer.

The great proportion of lime added to Rondelet's Stucco III gives rise to calcium carbonate, which increases resistance to humidity and delays setting, offering more time to do the decorative motives in relief [6]. The lowest coefficient of water absorption values are observed for Scagliola with glue and for mortars with sand, Cennin-i's Intonaco and Escaiola. In interior plasters the animal glue of Sca-gliola improves the impermeability of the surface by reducing the number and size of pores in the mortar [36], thereby protecting it from moisture. In Scagliola, the densification of the mortar, due to the filler effect of the ultramarine blue and rabbit glue, reduces capillarity absorption [4,32]. In fact, the ultramarine blue not only decreases the coefficient of water absorption, but also the rate of the water flow inside the gypsum mortar. This pigment decreases both the porosity and the pore interconnectivity in the Scagliola [4].

Cennini's Intonaco presents a good profile for the 2nd layer where the composition is outlined and then is covered by the Intonachino layer or Whitewash base. Escaiola presents positive characteristics to be applied as a decorative 3rd layer, appropriately protected against environmental agents like atmospheric water vapors. In the Escaiola and Stucco II the adsorption of cadmium onto silica of the sand, marble dust and limestone dust must contribute to decreasing the coefficient of water absorption; the selenite which is in the dark yellow cadmium is characterized by low porosity, and the addition of lime improved both the toughness and the low porosity of the mortars [36,38].

The low rate of water exchange between the different layers of the plasters and stuccoes facilitate the diffusion of water throughout the thickest layer of mortar, providing a decompression system [15]. The resistance to the diffusion of water of the Floating Coat and Stucco II is due to the use of limestone dust and marble dust, increasing the air content [25]. The drying test of gypsum-based mortars and of the formulated eco-mortar demonstrated that the

Table 7

Performance of the mortars, referred in technical literature (15th century to 19th century), historical and current: observed tendencies.

Measurement (unities) Gypsum-based mortarsa Gypsum-based mortarsb Lime-based mortarsc Lime-based mortarsd Lime-based mortarse

Min. Max. Aver. Min. Max. Aver. Min. Max. Aver. Min. Max. Aver. Min. Max. Aver.

BD (kg/m3) 754 1549 1152 600 1500 1050 1281 1663 1472 1300 - 61300 1394 1995 1695

Edyn (MPa) 1537 3163 2350 - - - 3538 7911 5725 2000f 5000f 610,000 3500f 874 7055 3965

CS (N/mm2) 1.3 2.5 2 2 6 4 1.2 2.2 1.7 0.4 0.6f 5 3f 2.5f 2.7 1.8f 1.6f 3 3.5 3.3

FS (N/mm2) 0.9 1.9 1.4 1 2 1.5 0.2 0.7 0.5 0.4 0.2f 0.8 0.7f 0.6 0.5f 1 2 1.5

CWA (kg/m2 min0 5)10-90 min 2.4 8.2 5.3 0.7 1.9 1.3 1.4 3.1 2.3 0.2 1.5f 0.4 1f 0.3 1.3f 0.1 0.9 0.5

P(%) 22.6 71.7 47.2 - - - 17.1 33.7 25.4 - - - 9.2 17.3 47.2 23 32.4

BD - bulk density; Edyn - dynamic E-Modulus; CS - compressive strength; FS - flexural strength; CWA - coefficient of water absorption; P - porosity. a Gypsum-based mortars referred in technical literature of Leitao (1896): at 90 days prismatic specimens (160 mm x 40 mm x 40 mm). b Current gypsum according to EN13279-1:2008.

c Lime-based mortars referred in technical literature from early 15th century, in Cennini, to 19th century, in Rondelet (1802) and Leitao (1896): at 90 days prismatic specimens (160 mm x 40 mm x 40 mm). d Current lime-based mortars as EN 998-1.

e Historical mortars referred by L'ANAH [40], Demelenne et al. (2010) [41], Veiga et al. [42]. f Renovation rendering and repointing, in Veiga [39].

O P (%) □ BD A CStr/P

"E CT"

Cennini's Stucco III Escaiola Stucco II Floating C. Setting C. Scagliola

Fig. 8. Compressive strength (CStr) according to porosity (P) and bulk density (BD), at 90 days. Legende: P (%); BD (Kg/m3 x 10~2); CStr (N/mm2 x 102).

water retention in the mortar is higher in Stucco II due to the fineness of limestone dust, and in Scagliola due to animal glue (Fig. 7).

3.5. Overall analyses

Table 7 compares the performance tendencies observed in the mortars prepared according to the technical literature from 15th century to 19th century, and in historical and current mortars [39-42]. This Table shows how the older recipes show a higher porosity and a lower strength. Table 7 and Fig. 8 show that the lowest strength and highest porosity seem to be correlated in the studied mortars.

4. Conclusions

Traditional and regional practices are sources of knowledge and skills which can be reactivated in contemporary building construction and in historical building restoration and repair. Hence, is contributing to diversifying and optimizing the design of renders and plasters of lime-or-gypsum-based mortars. In comparison with present day industrial mortars, the ancient mortars reproduced in this research, seem to be heavier, with a lower mechanical strength, as well as a lower resistance to water. However, their longevity and durability indicate the need for determining their chemical, mechanical and hygroscopic profile.

Limestone dust and marble powder are industrial wastes which can successfully (in terms of performance) be utilized as fillers in lime-based mortars.

In this study the lime-based compositions with marble dust or limestone fines, similar to those described in ancient treatises like Leitao's Escaiola and Stucco II, for application in interior decorative stuccoes, and Rondelet's Stucco III and Leitao's Floating Coat for interior and decorative stuccoes and plasters, give compressive strength and water absorption results within the specification parameters [31,32]. Replacement of river sand by industrial waste, lime stone fines, in Stucco II 100%, improved the compressive strength as indicated in the Leitao's Escaiola control mortar. Cadmium is chemically stable and highly insoluble hence can be introduced as pigment in the Escaiola and Stucco II. Ultramarine blue has a high capacity of gypsum fixation, is highly sulfate-resistant and reduced the expansion of the specimens at 90 days of exposure to sulfate sodium, improving the durability and compressive strength of the gypsum-based plaster with glue water, Scagliola. Both the glue water and the ultramarine blue reduced capillarity absorption of Scagliola. The glue water decreases the compressive strength but increases the flexural strength.

It was found that gypsum-based mortars (Leitao's Floating Coat, Setting Coat and Scagliola) have better performance in interior plasters, which is further promoted by the high dissolution of gypsum and the plaster's high capillary absorption (Leitao's Setting Coat). Gypsum makes the mortar stiffer, thereby reducing the tensile stresses that prevent the decorative last layer from cracking, which is a positive characteristic, especially for covering flexible supports. So, the gypsum mortars are suitable for wooden plaster frameworks, like ceilings or partition walls. Floating Coat and Stucco II, having a lower rate of water absorption than Setting Coat and Rondelet's Stucco III, are suitable for the 2nd layer of the plaster or stucco covering system, and Setting Coat and Rondelet's Stucco III are suitable for the 3rd and finishing layer. Nevertheless, Stucco II can be applied in the outermost layer if the precedent layer is a non-hydraulic lime putty binding system mixed with calcareous sand, for example.

The mortars based on lime putty with sand (Cennini's Intonaco, Leitao's Escaiola), due to their low coefficient of water absorption and shrinkage, present good resistance to humidity and so are appropriate for surfaces which demand greater durability. However, Cennini's Intonaco has a low mechanical strength, and needs

to be covered by a layer with better mechanical performance, prepared with slaked lime and marble cream and dust (¡ntonachino).

The lime-based composition, with marble dust or limestone filler or fines, gives the best mechanical results. The 100% substitution of sand by limestone fines (Stucco ¡¡) increases its compressive strength by 19.5% and the flexural strength by 16.1%, while decreasing the resistance to humidity by 18.6%. Renders and plasters with marble dust (Rondelet's Stucco ¡¡¡, Leitaos Escaiola) have a better aesthetic effect, moderate values of capillary absorption and a high water resistance, properties that are important for their strength and durability. Rondelet's Stucco ¡¡¡, Leitao's Escaiola, Stucco ¡¡, Setting Coat and Scagliola have higher values for compressive or flexural strength, a necessary property of the stucco or plaster for the finishing layer in comparison with the precedent layers.

Stuccoes and plasters reproduced based on ancient treatises are mainly suited for renovation plasters/rendering systems and preservation of historical buildings due to their porosity for vapor permeability and their significantly reduced capillary conductivity. The present research can contribute to development in design, optimization, and manufacturing of a new generation of building materials.

Acknowledgments

The authors thank Dr. Rosario Veiga the Coordinator of the Renders and Insulation Nucleus of Building Department of LNEC (National Laboratory of Civil Engineering, in Lisbon), and Eng. Nelson Duarte of Tests, Corrosion and Materials Laboratory of Pedro Nunes Institute (IPNled&mat, in Coimbra) for providing the necessary facilities for this research work.

References

[1] Agarwall SK, Gulati D. Utilization of industrial wastes and unprocessed micro-fillers for making cost effective mortars. Constr Build Mater 2006;2:999-1004.

[2] Al-Akhras N, Ababneth A, Alaraji WA. Using burnt stone slurry in mortar mixes. Constr Build Mater 2010;24:2658-63.

[3] Machida M, Fotoohi B, Amamo Y, Mercier L. Cadmium(II) and lead(II) adsorption onto hetero-atom functional mesoporous silica and activated carbon. Appl Surf Sci 2012;258:7389-94.

[4] Giraldo C, Tobon JI, Baena OJR. Ultramarine blue pigment: a non-conventional pozzolana. Constr Build Mater 2012;36:305-10.

[5] Papayianni I, Stefanidou M, Christodoulou S. Influence of sand fines on the mechanical and physical properties of lime-based renders and plasters. In: 2nd Historic mortars conference HMC2010 and RILEM TC 203-RHM final workshop. Prague: RILEM; 2010. p. 1135-42.

[6] Bueno AG, Florez VJM. The Nasrib plasterwork at ''qubba Dar al-Manjara l-Kubra ''in Granada: characterization of material and techniques, in. J Cult Heritage 2004;5:75-89.

[7] Cennini CA. Early 15th century. Translated by Daniel V. Thomson, Jr. New York Dover Publications; 1933. ''Il Libro del'Art". 3th Section. Chap. LXVII. Internet. <http://www.noteacess.com/texts/Cennini/3htm>, (accessed 29.08.2000).

[8] Mottez V. Le Livre de l'art ou traité de la peinture par Cennino Cennini. Paris: F. de Nobele Libraire; 1982.

[9] Robison GT. Plasterwork generally: a glimpse of its history. In: William Millar, George Bankart, editors. In plastering plain & decorative. UK, USA: Donhead Publishing Ltd.; 2009. p. 1-28.

[10] Ortiz Sanz JF. Los Quatro Libros de Arquitectura de Andres Palladio, Vicentino. Madrid: Imprenta Real. 1797; Livro I, Cap. V.

[11] Torraca G. Lectures on materials science for architectural conservation. The Getty Conservation Institute, Los Angeles, United States; 2009. p. 51,55,56. ISBN: 978-09827668-3-5.

[12] Ouezdou MB, Hajjem T, Mensi R. Characterization of waterproof-covering mortars on Ottoman Monuments of Ghar El Melh (Tunisia). Constr Build Mater 2009;23:423-33.

[13] Dheilly RM, Tudo J, Sebambi Y, Quèneudc M. Influence of storage conditions on the carbonation of powdered Ca(OH)2. Constr Build Mater 2002;16:155-61.

[14] Marinowits C, Neuwald-Bur C, Pfeifer M. Historic documents in understanding and evaluating of historic lime mortars.In: 2nd Historic mortars conference HMC2010 and RILEM TC 203-RHM final workshop, 22-24 September 2010, Prague, Czech Republic; p. 267-74.

[15] D'Orazio M, Stazi A. Dinamic of moister transfer in ancient plasters. J Cult Heritage 2006;7:116-22.

[16] Rampazzi L, Rizzo B, Colombo C, Conti C, Realini M, Bartoluci U, et al. The stucco decorations from St. Lorenzo in Laino (Como, Italy): the materials and the techniques employed by the "Magistri Comacini". Anal Chim Acta 2008;630:91-100.

[17] Rondelet JB. Traité Théorique et Pratique de l'Art de Bâtir. Paris: Chez l'Auteur, Enclos du Panthéon; 1802.

[18] Salavessa E, Shahsavandi A, Torgal F, Jalali S. Causes of decay of the Aveleiras House decorative plasters (in Torre de Moncorvo). In: 2nd Historic mortars conference HMC2010 and RILEM TC 203-RHM final workshop, Prague; 2010. p. 683-90.

[19] Leitào LA. Elementary Course of Construction, elaborated according to the central school of the same arm program (Curso Elementar de Construçào, elaborado segundo o programa da escolar central da mesma arma). Lisboa: Imprensa Nacional; 1896.

[20] Millar W, Bankart G. Fibrous plaster. In: William Millar, George Bankart, editors. In plastering plain & decorative. UK, USA: Donhead Publishing Ltd.; 2009. p. 241-73.

[21] Gil M. Terras coradas em Portugal: da fonte de matéria-prima à produçào de pigmento terra nacionais (sécs. XVIII ao XX). 9" Seminârio Ibero-Americano de Arquitectura e construçào com Terra, Argumentum; 2010. p. 20-4.

[22] Cruz AJ. The artists colours: history and science of pigments used in painting (As Cores dos Artistas: Histôria e Ciência dos Pigmentos utilizados na Pintura) "Omniciênia", no. 5, Lisboa; 2004. p. 11 and 12.

[23] Coelho AZ, Torgal FP, Jalali S. Lime in construction (A Cal na Construçào). Guimaràes: Ed. TecMinho, Universidade do Minho; 2009.

[24] CEN-EN 1015-3:1999. Methods of test for mortar for masonry - Part 3: determination of consistence of fresh mortar (by flow table). CEN, Brussels; 1999.

[25] Topçu IB, Bilir T, Uygunoglu. Effect of waste marble dust content as filler on properties of self-compacting concrete. Constr Build Mater 2009;23:1947-53.

[26] Corinaldesi V, Moriconi G, Naik TR. Characterization of marble powder for it use in mortar and concrete. Constr Build Mater 2010;24:113-7.

[27] CEN-EN 1015-11:1998. Methods of test for mortar for masonry - Part 11: determination of flexural and compressive strength of hardened mortar. CEN, Brussels; 1999.

[28] CEN-EN 13279-2:2004. Gypsum binders and gypsum plasters - Part 2: test methods. CEN, Brussels; 2004.

[29] CEN-EN 1015-18:2002. Methods of test for mortar for masonry - Part 18: determination of water absorption coefficient due to capillary action of hardened mortar. CEN, Brussels; 2002.

[30] Ventolà L, Vendrell M, Giraldez P, Merino L. Traditional organic additives improve lime mortars: new old materials for restauration and building natural stone fabrics. Constr Build Mater 2011;25:3313-8.

[31] Quagliarini E, Lenci S, Seri E. On the damage of frescoes and stuccoes on the lower surface of historical flat suspended light vaults. J Cult Heritage 2011. http://dx.doi.org/10.1016/J.culher.2011.11.00.

[32] Faria P, Henriques F, Rato V. Comparative evaluation of lime mortars for architectural conservation. J Cult Heritage 2008;9:338-46.

[33] Kovler K, Frosting y. On the problem of cracking in plaster layers. Constr Build Mater 1998;12:251-8.

[34] Lawrence RMH, Mays TJ, Walker P, D'Ayala D. Determination of carbonation profiles in non-hydraulic lime mortars using thermogravimetric analysis. Thermochim Acta 2006;444:179-89.

[35] Silveira PM, Veiga MR, Brito J. Gypsum coatings in ancient buildings. Constr Build Mater 2007;21:126-31.

[36] Machida M, Fotoohi B, Yoshimasa A, Mercier L. Cadmium(II) and lead(II) adsorption onto hétero-atom functional mesoporous silica and activated carbon. Appl Surf Sci 2012;258:7389-94.

[37] Genestar C, Pons C. Ancient covering plaster mortars from several convents and Islamic ad Gothic palaces in Palma de Mallorca (Spain). Analytical characterization. J Cult Heritage 2003;4:291-8.

[38] Franzoni E, Sandrolini F, Baldazzi L. Characterization of gypsum-selenite plasters from historic buildings in Emilia - Romagna Region (Italy). In: 2nd Historic mortars conference HMC2010 and RILEM TC 203-RHM final workshop, Prague, Czech Republic; 2010. p. 157-164.

[39] Veiga MR. Lime-based mortars in conservation of old buildings (Argamassas de cal na conservaçào de edificios antigos).In: Engineering conferences on reabilitation of buildings. Lisbon: LNEC; 2006.

[40] L'ANAH. Etude performancielle d'enduits anciens/rapport final. La Chaux et les Mortiers: natures, proprietés, traitements. In: ICOMOS scientific symposium on lime mortars, Paris; 1995. p. 14.

[41] Demelenne M, Dagrain F, Scaillet JC. Development of a methodology for characterization ofhistorical mortars in the Walloon Region (Belgium). In: 2nd historic mortars conference HMC2010 and RILEM TC 203-RHM final workshop, Prague, Czech Republic; 2010. p. 99-109.

[42] Veiga R, Magalhàes A, Bokan-Bosilikov V. Capillarity tests on historic mortar samples extracted from site/methodology and compared results. Lisbon: LNEC; 2004. p. 4-5.