Monument Future

Introduction

Stone materials involved in fires may be affected by a variety of phenomena, which can change their original properties and related performance in buildings. Depending on the temperature, both chemical and mineralogical trasformations may occur and lead to stone colour changes (Kompaníková et al. 2014). In addition, volume variations due to phase transitions (Calia et al, 2015), different thermal expansion of adjacent minerals (Vázquez et al., 2015) or strongly anisotropic thermal properties of some minerals (e. g. calcite) (Siegesmund et al., 2000) may affect the stone microstruture and lead to microfissuring, which may increase the stone susceptivity to weathering and compromise its load-bearing performance (Sippel et al., 2007). Thermally induced effects on the microstructure strongly depend on the inherent stone characteristics and have a high incidence on low porosity materials, due to the dense packing of crystals and grains (Yavuz et al., 2010). Thermal behaviour has been studied for a variety of stones, which mainly include compact materials (Martinho et al., 2018), while poor literature deals with heating damage on porous stones (Gomez-Heras et al., 2006; Brotóns et al., 2013; 78Franzoni et al., 2013). In this paper we present the results of a case study where the effect of a fire on highly porous calcarenites were assessed by using integrated investigation techniques.

Material

The study was carried out on the stone materials used within the masonry of the ACAIT (Azienda Cooperativa Agricola Industriale del Capo di Leuca) industrial building (Fig. 1). The building was a factory for the processing of the tobacco, in the province of Lecce (Southern Italy). It was built in the early 1900 and dismissed at the end of the 1980s. The factory is a remarkable example of the industrial archaeological heritage relating to the tobacco manufacturing, a flourishing activity in which Puglia region had a leading role on a national basis in the past century.

The original building developed on the ground floor only and it has undergone several enlargements over the years. Like all industrial buildings, it has a simple and modular layout, composed of large rooms with corner-vaulted (“volta a spigolo”) roof, typical of the local tradition, load-bearing pillars and masonry walls with regular local limestone ashlars.

Figure 1: The building of the tobacco factory ACAIT affected by a partial collapse.

Figure 2: Detail of the stone from the collapsed vaults, which shows a yellow-beige-color (Y) passing to a reddish one (R) across the thickness of the masonry unit.

In 2018, after a strong rainstorm, part of the structure collapsed (Fig. 2), involving in a first time one vault and part of the two adjacent ones and in a second time (about 7 days later) the remaining part (five vaults) of the room. The collapse evidenced the inner structure of the vaults and external walls. The latter were two leafs walls without horizontal connections, filled with incoherent material (pieces of rocks and debris).

The vault collapse revealed traces of an historic fire, which were hidden by the presence of plasters, in the form of fumes and a diffuse reddish color across the thickness of the masonry units up to some centimeters from the surface (Fig. 2).

In the framework of a diagnostic activity supporting a restoration project in view of a building reuse, the study of the fire damage on the stone was undertaken.

Methods

Collapsed blocks measuring 21x20x50 cm were taken from the site and samples were obtained from both the unaltered and altered portions, having yellow-beige (Y) and reddish colour (R), respectively (Fig. 2).

The following analyses and tests were performed.

— Thin-section samples were observed in plane-polarised and cross-polarised transmitted light by means of an optical microscope (Eclipse LW100 Nikon) at magnifications of 50x e 100x.

— X-Ray Diffraction analyses (XRD) were performed on both the whole rock and insoluble residue. The insoluble residue was separated by a chemical attack of the grinded stone with HCl-3N in order to remove the carbonates by dissolution. A Philips 1742 diffractometer (APD – 3.6j version) was used for the analyses (CuKα, 40 kV, 20 mA, 2ϑ step size of 0.02 °, counting time 1.25 s, scan interval between 3 ° and 60 °). The diffraction data 79were processed with a X’Pert software — Philips Analytical.

— Simultaneous thermal analyses by Differential Scanning Calorimetry and Thermogravimetry (DSC-TG). A Netzsch STA 449 F3 Jupiter^® was used. Both samples of the whole rock and of the insoluble residue were analyzed. Approximately 25 mg of each powder sample were heated in air from the ambient temperature to 1,000 °C, at a heating rate of 10 °/min.

— Colour changes were recorded by colorimetric measurements. These were taken by light absorption in diffuse reflection using a Konica Minolta CM700d spectrophotometer. They were carried out with a D65 illuminant and under a 10 ° standard observer. L*, a* and b* colour coordinates in the CIELab system were measured and the colour variation (ΔE*) was calculated.

— Measurements of bulk density, porosity accessible to water and water absorption amounts of the stone samples were performed through saturation and buoyancy techniques, following the ISRM recommendation [ISRM, 1981].

— Ultrasonic Pulse Velocities (UPVs) were measured on specimens (cubes 70 mm side obtained from masonry blocks coming from the site) after drying at 70 °C, according to ASTM D2845-05 (ASTM 2005). In particular, three visible unaltered (Y) and three colored specimens (R) were taken from the collapsed portion of the building. Velocities were measured by direct transmission method using a TDAS 16 (Boviar) instrument and probes with a frequency of 55 kHz. They were recorded in each direction (x, y, z) of the cubic specimens and expressed as mean values.

— Compressive strength tests were performed according to UNI EN 772-1 (UNI 2011) on the same specimens used for UPV test, after drying at 70 °C. A universal testing machine (Metrocom Engineering spa), with a load capacity of 200 kN and a speed of 0.2 mm/min, was used for the test.

Results and Discussion

The petrographic characteristics, as observed by optical microscopy under polarized transmitted light (Fig. 3a), show that the investigated stone is a medium grainstone. It is almost exclusively made of calcareous fossil remains, which mainly consist of coralline algae and, at lower extents, of benthic foraminifera, echinoids, bivalves and bryozoans. The average dimensions of the bioclasts fall between 0.3 and 0.4 mm with a maximum size of 0.6 mm. The stone contains sporadic quartz and feldspar crystals. The micrite is nearly absent and the cement is made of calcite, with a texture varying from microsparitic to sparitic type. It is in poor amount and fills only partially the interparticle porosity, which results very high. At large extents the cement is in the form of a thin level surrounding the grain borders. In some areas it is present in larger spots and exhibits a well-developed sparitic texture.

Figure 3: Microscopic features of the stone (N//). a: yellow-beige level; b: red level.

No damage in the form of microfissuring affecting the stone microstructure was observed in the red portions, compared to the yellow-beige ones (Fig. 3b). On the contrary, there was an increase of red pigmented bioclasts and iron rich agglomerations, which may be related to an effect of the high temperature on the iron rich components.

80

Figure 4: XRD spectra of the whole rock (on the top) and insoluble residue (on the bottom) from the yellow (Y) and red (R) levels.

The mineralogical composition, as determined by XRD analyses of the whole stone samples coming from the levels having the different colours (Fig. 4, top) does not show any differences. In all cases, almost exclusively calcite was detected. A diffraction peak at low angles was visible, as relating to the presence of clay minerals.

To detect the presence of other minerals, masqued by the preponderant CaCO₃ in the whole stone composition, the insoluble residue, after removing all carbonates by dissolution in HCl-3N, was also analysed. The XRD powder patterns of the insoluble residue obtained from the yellow-beige and red levels in the stone after this chemical attack are reported in Fig. 4, bottom. Different mineralogical compositions were found. The presence of quartz, goethite, along with some feldspars and clay minerals was detected in the yellow-beige level. Goethite was absent in the red level, instead hematite was detected. The transformation of goethite to hematite comes from a dehydroxylation process. Such a transition takes place at temperature of 300 °C (Földvári 2011).

Results of the simultaneous TG and DSC analyses performed on the whole stone from the yellowbeige and rel levels are illustrated in Fig. 5.

TG curves well recorded the calcite decomposition between 670 °C and 840 °C with a mass loss of about 40 %.

Figure 5: TG/DSC curves of the whole rock from the yellow (Y) and red (R) levels.

Figure 6: TG, DTG and DSC curves for the insoluble residue from the yellow (Y) and red (R) levels.

Calcite decomposition in the same range of temperature is evident in the DSC curves through the 81presence of an endothermic peak. A slight endothermic peak at about 80° C is also present, along with a larger one at 550 °C. Both are better shown in the DSC curves of the insoluble residue. TG/DTG curves (Fig. 6) of the insoluble fraction from the yellow-beige levels show a first mass reduction, with a peak in the DTG curve at 84 °C. In the R sample this mass loss is shifted at 95 °C and it is less pronounced. These thermal variations are consistent with dehydration due to the evaporation of the adsorptively bound water from the specimen (Földvári 2011). A second mass loss with a peak in the DTG curve at 287 °C is observed in the sample from the yellow-beige stone, which can be attributed to the goethite dehydroxylation (Földvári 2011). This peak is absent in the DTG curve of the R sample according to the XRD findings, which did not detect goethite in this sample, but hematite as a product of its transformation. For temperatures higher than 400 °C, the pattern of the DSC curve evidences an endothermic-exothermic process. It corresponds to a solid-phase structural decomposition of organic matter and clay minerals, which is more pronounced in the Y sample compared to the R one. It is followed by a crystallization of new phases, whose evidence is given by a subsequent exothermic bump.

Colour changes, bulk density, porosity accessible to water and water absorption measured for the stone from the yellow-beige and red levels within the blocks are reported in Table 1.

A strong colour variation was recorded in the red level, which may be attributed to the transition of goethite to hematite detected through the mineralogical and thermal analyses. High porosity and water absorption, as well, were measured in the yellow level and not significant decreases of 6 % and 5 %, respectively, were measured in the red level. Also the bulk density showed a slight reduction, namely 3 %. These variations are comparable with those reported for limestones with high porosity and notably lower than the decreases recorded for compact stones (Gomez-Heras et al., Brotons et al. 2013). They were in the range of variability of the measurements, thus they could be due to the intrinsic stone heterogeneity.

However, a decrease of UPV in the samples from the red level was recorded. It was 21 % (Table 1).

UPV decrease as an effect of the heating is reported in the literature (Yavuz et al. 2010, Andriani 2014), although at entities depending on the temperature and stone structure, as well. It mainly relates to a thermal micro fissuring, which causes the reduction of the propagation velocities.

The microfissuring detected by the UPV test slightly affected the above mentioned physical parameters measured by stauration and buoyancy techniques. This finding suggests that the recorded decreases of the wave velocities may be relevant to the generation of a microporosity which has no effect on the water penetration, as reported in previous studies (Franzoni et al., 2013; Freire-Lista et al., 2016). Microfissuring recorded by UPV had a negligible effect also on the mechanical performance of the stone. Very close values of the compressive strength were measured in both yellow and red levels, corresponding to a strenght loss of 4 % in the discoloured level (Table 1). Similar entities of decrease have been recorded for porous limestones by Franzoni et al., 2013.

Conclusions

Macroscopic evidences of a fire in the calcarenites employed in an historic building were confirmed by mineralogical changes, which reflects 82on strong color changes. In particular, the change from yellow-beige to reddish color of the stone is consistent with the thermally induced transformation of goethite to hematite. This transition phase indicates that temperatures around 300 °C were reached in the red stone levels during the fire. Effects on the stone microstructure were not visible under optical microscope. Nor the measurement of physical properties showed meaningful variations in this regard. On the contrary, UPV detected a decrease of the propagation velocities, which probably denotes a stone microfissuring. Nonetheless, its entity did not compromise the mechanical resistance of the stone, which remained almost unchanged. High porosity may account for a slight microstructural damage recorded for the investigated calcarenite, where pores likely behave as free spaces for expansion of calcite grain preventing in this way an extensive damage.

Table 1: Color change (ΔE), bulk density (γb), porosity accessible to water (P), water absorption (WA), ultrasonic pulse velocity (UPV) and uniaxial compressive strength (UCS) measured for the stone from the yellow-beige and red levels.

References

Andriani G. F., Germinario L. 2014. Thermal decay of carbonate dimension stones: fabric, physical and mechanical changes. Environ Earth Science 72: 2523–2539.

ASTM D2845-05 2005. Standard test method for laboratory determination of pulse velocities and ultrasonic elastic constants of rock. American Society for Testing Materials.

Brotóns V., Tomás R., Ivorra S., Alarcón J. C. 2013. Temperature influence on the physical and mechanical properties of a porous rock: San Julian’s calcarenite. Engineering Geology 167: 117–127.

Calia A., Colangiuli D., Lettieri M., Quarta G., Masieri M. 2016. Microscopic techniques and a multi-analytical approach to study the fire damage of the painted stuccoes from the Petruzzelli Theatre (Bari, Southern Italy). Microchemical Journal. 126: 42–53.

Földvári M. 2011. Handbook of thermogravimetric system of minerals and its use in geological practice. In: Geological institute of Hungary. Budapest. isbn 978-963-671-288-4.

Franzoni E., Sassoni E., Scherer G.W, Naidu S. 2013. Artificial weathering of stone by heating. Journal of Cultural Heritage. 14: 85–93.

Freire-Lista D. M., Fort R., and Varas-Muriel M. J. 2016. Thermal stress-induced microcracking in building granite. Engineering Geology 206:83–93.

Gomez-Heras M., Alvarez de Buergo M., Fort R., Hajpál M., Török A., Varas M. J. 2006. Evolution of porosity in Hungarian building stones after simulated burning. In: Heritage Weathering and Conservation HWC-2006, Taylor & Francis, Rotterdam, 513–519.

ISRM. Rock characterization testing and monitoring. In: Brown ET, editor. ISRM suggested methods, Oxford:Pergamon Press; 1981.p. 135–40.

Kompaníková Z., Gomez-Heras M., Michňová J., Durmeková T., Vlčko J. 2014. Sandstone alterations triggered by fire-related temperatures. Environmental Earth Science 72: 2569–2581.

Martinho E., Dionísio A. 2018. Assessment Techniques for Studying the Effects of Fire on Stone Materials: A Literature Review. International Journal of Architectural Heritage. 00 1–25.

Siegesmund S., Ullemeyer K., Weiss T., Tschegg E. K. 2000. Physical weathering of marbles caused by anisotropic thermal expansion, Int. J. Earth Sci. 89: 170–182. doi:10.1007/s005310050324.

Sippel J., Siegesmund S., Weiss T., Nitsch K. H., Korzen M. 2007. In: Pikryl, R. & Smith, B. J. (eds) Building Stone Decay: From Diagnosis to Conservation. Geological Society, London, Special Publications 271: 139–151.

UNI EN 772-1 2011. Methods of test for masonry units – Part 1: Determination of compressive strength.

Yavuz H., Demirdag S., Caran S. 2010. Thermal effect on the physical properties of carbonate rocks. International Journal of Rock Mechanics & MiningSciences 47: 94–103.

Vázquez P., Shushakova V., Gómez-Heras M. 2015. Influence of mineralogy on granite decay induced by temperature increase: Experimental observations and stress simulation. Eng. Geol. 189: 58–67. doi:10.1016/j.enggeo.2015.01.026

83

VARIATIONS OF CHARACTERISTICS OF SANDSTONE SUBJECTED TO WEATHERING AND CONSERVATION INTERVENTIONS

Miloš Drdácký¹, Dita Frankeová², Zuzana Slížková²

IN: SIEGESMUND, S. & MIDDENDORF, B. (EDS.): MONUMENT FUTURE: DECAY AND CONSERVATION OF STONE.

– PROCEEDINGS OF THE 14TH INTERNATIONAL CONGRESS ON THE DETERIORATION AND CONSERVATION OF STONE –

VOLUME I AND VOLUME II. MITTELDEUTSCHER VERLAG 2020.

1 Institute of Theoretical and Applied Mechanics of the Czech Academy of Sciences, Department of Heritage Science, Prosecká 76, 190 00 Praha 9, Czech Republic, drdacky@itam.cas.cz

2 Institute of Theoretical and Applied Mechanics of the Czech Academy of Sciences, Department of Material Research, Prosecká 76, 190 00 Praha 9, Czech Republic

Abstract

The paper presents selected results of a comprehensive study of characteristics and behavior of seven typical sandstone types used for historic buildings in the Czech Republic, mostly in Prague. Stone characteristics were studied on materials affected by various historic environmental impacts and conditions generated by previous interventions. From seven types of sandstone were prepared nine series of test specimens which included chemically deteriorated surface layers (crust), cleaned surface layers, and virgin material from the stone core. The above-described sets were manufactured without any consolidation treatment as well as in two further sets consolidated with two agents, namely Funcosil 100 and 300, based on the silicic acid ester. The test specimens were cut from damaged sandstone blocks, which were extracted from a masonry rail of the Charles Bridge in Prague before replacement with new elements. The results supplied data for comparing the efficiency of the consolidation treatment with silicic acid ester products in relation to three pre-treatment stone conditions, as well as to the type of sandstone cementation (mostly a kaolin or silica, rarely goethite cementation). In the paper, the most important results and conclusions taken from the tests and their comparison are discussed.

Introduction

Charles bridge in Prague has been subjected to various types of deterioration or damaging actions during its nearly eight hundred years history. As a result, some parts were substantially repaired using different types of sandstone available at the given periods. The stone materials exhibit different characteristics decisive for the application of efficient conservation or maintenance technologies. Therefore, a detailed investigation programme has been launched in order to provide restorers with reliable data on material characteristics as well as the response to a selected pilot consolidation treatment.