The Effect of Fire on Building Materials

heritage

Article The Eﬀect of Fire on Building Materials: The Case-Study of the Varnakova Monastery Cells in Central Greece

Ekaterini T. Delegou 1,*, Maria Apostolopoulou 1, Ioanna Ntoutsi 1, Marina Thoma 1, Vasileios Keramidas 1, Christos Papatrechas 2, George Economou 2 and Antonia Moropoulou 1,*

1 Laboratory of Materials Science & Engineering, School of Chemical Engineering, National Technical University of Athens (NTUA), 9 Iroon Polytechniou Str., Zografou Campus, 15780 Athens, Greece; [email protected] (M.A.); [email protected] (I.N.); [email protected] (M.T.); [email protected] (V.K.) 2 Institute of Geology and Mineral Exploration (I.G.M.E.), Mineralogy and Petrology Department, 1 Spyrou Loui, Acharnes, 13677 Eastern Attica, Greece; [email protected] (C.P.); [email protected] (G.E.) * Correspondence: [email protected] (E.T.D.); [email protected] (A.M.); Tel.: +30-2107-723-075 (E.T.D.); +30-2107-723-276 (A.M.) Received: 21 February 2019; Accepted: 17 April 2019; Published: 20 April 2019

Abstract: The evaluation of the fire impact on building materials is of great scientific and socio-economic importance since fire can result in materials’ chemical and mechanical alterations, which leads to structural stability problems of historical and/or modern construction. This highly increases the cost of rehabilitation interventions. The case study of the Byzantine Monastery of Panagia (Virgin Mary) Varnakova is an example of the fire effect on both historical and newer stone masonries. The Varnakova Monastery is a typical 19th century monastic complex and, during its long history, it has undergone multiple reconstructions after major catastrophic events that have taken place due to its strategic geographical position and its financial and spiritual significance for the region. The last big-scale renovation of the Monastery was conducted between the years 1992 to 2014. However, in January 2017, a devastating fire destroyed the largest part of the monastic cells’ quarter. In this work, a diagnostic study of the different construction phases’ materials comprising the masonries of the monastery cells in their present state is presented. The examination of a series of samples through analytical techniques, such as optical microscopy, X-ray diffraction, thermal analysis, and total immersion tests, along with the use of non-destructive techniques in situ, such as Infra Red Thermography, Digital Microscopy, and Schmidt Hammer Rebound tests, shed light on the preservation state and on the decay of the diverse building materials. In addition, the impact of the fire on their properties was investigated. The results reveal the diversity of the materials used in the historical masonries throughout the centuries, while the combination of analytical and non-destructive techniques demonstrates the damages induced by the fire.

Keywords: stone; mortar; masonry; ﬁre; NDT; IRT; SHR; microscopy; petrography; XRD; TG/DTA

1. Introduction In recent years, more research has focused on the effect of high temperatures on building materials [1–16]. The development of high temperatures affects the various characteristics of a material. Petrographic, mineralogical, chemical, physical, and mechanical properties may alter, depending on the temperatures developed, the characteristics of the material, the timeframe of the high temperature event, and the manner of cooling. The assessment of the effect of high temperatures on the building

Heritage 2019, 2, 1233–1259; doi:10.3390/heritage2020080 www.mdpi.com/journal/heritage Heritage 2019, 2 1234 materials of a structure, especially after a catastrophic event, such as a fire, is crucial in order to evaluate the safety level of its damaged state and plan appropriate repair measures [10]. This assessment is even more important in the case of monumental structures, where the historic building materials are “precious” and should, if possible, be replaced to the minimum extent. The effect of high temperature on a building material is a complex mechanism and various material properties must be examined in order to obtain a holistic view of a fire effect. Visual inspection provides useful information regarding the effect of a fire event, since many damages and decay patterns are macroscopically manifested, including cracking, surface crazing, deflection, delamination, color changes, and smoke damage [12,13,15,16]. Petrographic analysis assists not only in the characterization of a stone type, which reveals composition, texture, binder type, and micro-morphology, but in the evaluation of the effect of high temperature as well, by examining the alterations observed in these characteristics [2,4–7,10]. A temperature increase induces inter-granular and intra-granular effects on a material while inter-granular effects are connected to the development of microcracks within the matrix of the material and are usually the result of different characteristics of the material’s constituents. Intra-granular effects manifest through the development of microcracks within the grains themselves [1]. Inherent anisotropy of materials is an example of degradation in the case of thermal loads. For example, inherent anisotropy of calcite grains, when heated, is considered to be the main agent for marble and limestones disintegration [11,13]. Mineralogical analysis reveals the alteration of mineralogical compounds. Thus, the absence or decrease of a compound after the fire event is indicative that the fire reached a temperature at which the compound’s structure irreversibly collapses and/or is transformed into an amorphous compound that is no longer detectable [2]. Chemical analysis can reveal alterations in the composition of the material, which is also connected to transformations that take place during the fire effect. These alterations include oxidation, dehydration, de-hydroxylation, calcination, and other processes [6,14]. These processes may infer matrix alterations of a building material and/or color changes, such as reddening (e.g., when ferrous minerals are included in the material). Thus, for example, the clay and iron oxides-hydroxides content within a stone or aggregates of a mortar may be a determinant for a series of transformations induced by heat [16]. Microstructural analysis is of paramount importance since the alteration of microstructural characteristics affects other properties, such as the behavior of a material under hygro-thermic loads and compressive strength [1–7,9]. Porosity usually increases when a material is subjected to high temperatures [1,3], while density decreases due to material loss and compound decomposition, dehydroxilation, and dehydration processes. Mechanical characteristics are crucial to structural stability. Compressive strength usually decreases due to the development of microcracks and/or the increase of porosity [8,9]. The effect of fire and the consequent alterations induced to a material’s characteristics is clearly dependent on the material itself. In the case of stones, each lithotype presents a different behavior to fire-attack, depending on pertrographic characteristics, chemical composition, porosity, anisotropy of its components, etc. For example, granites are more prone to crack when thermally shocked, in comparison with other softer stones [5]. In quartz-rich sandstones, due to the open porosity they exhibit, as well as due to their calcitic matrix (which undergoes a parallel calcining process), micro-strains due to quartz expansion (when it is transformed from α β at ~575 C), are absorbed by the matrix [1]. Thus, → ◦ knowledge of the type of stone and parallel evaluation of alterations studied in related research and practice can assist in predicting the alterations induced in a certain lithotype in the event of a fire [12]. In terms of physico-mechanical and structural stability of the monument, it is crucial to assess the impact of fire damage. In an extent to the above discussion, the indirect consequences to the whole masonry need to be taken under consideration, during a diagnostic study. Another important aspect to be explored is the variability of the fire impact parameters on the different lithotypes and parts of the masonry depending on the diverse temperatures, duration of the event, or other synergic factors (e.g., precedent defaults of a masonry block or an isolated stone). Mortars, depending on the mortar type, Heritage 2019, 2 1235 Heritage2019, 2 FOR PEER REVIEW 3 diffusivitymay be more and or high less resistantincombustibility to a fire. event.However For, example,the decrease concrete in its presentsload-bearing low thermalcapability diff andusivity its “explosiveand high incombustibility. spalling” due to However,the fire action the decrease may compromise in its load-bearing these assets, capability especially and when its “explosive the fire temperaturespalling” due exceeds to the fire300 action°C [15,16 may]. Taking compromise into account these assets, all the especially above, the when fire theimpact fire temperatureneeds to be addressedexceeds 300 initially◦C[15, 16 in]. terms Taking of into building account materials all the above,. However, the fire itimpact aims needsto a holistic to be addressed perspective initially by studyingin terms ofthe building structure materials. as a whole. However, it aims to a holistic perspective by studying the structure as a whole.Within the above framework, this study is an attempt to investigate the effect of the 2017 fire eventWithin on the theVarnakova above framework, Monastery this building study is materials an attempt and to investigatethe masonries the etheyffect comprise of the 2017. Hereby fire event, the on preliminarythe Varnakova results Monastery of a diagnostic building materials study through and the an masonries in situ non they-destructive comprise. Hereby, testing theand preliminary an in-lab applicationresults of a of diagnostic analytical study techniques through on an selected in situ samples non-destructive are presented, testing which and an aim in-lab to understand application the of effectanalytical of the techniques fire on the on different selected sampleslithotypes are and presented, mortars which comprising aim to understandthe masonries the eofff ectthis of historic the fire structure.on the diff erent lithotypes and mortars comprising the masonries of this historic structure.

2. Materials and Methods 2. Materials and Methods 2.1. The Varnakova Monastery 2.1. The Varnakova Monastery The Byzantine Monastery of Panagia (Virgin Mary) Varnakova is located on a mountainous The Byzantine Monastery of Panagia (Virgin Mary) Varnakova is located on a mountainous landscape of Southwest Central Greece, at an altitude of 800 m (Figure1a). Its general structure follows landscape of Southwest Central Greece, at an altitude of 800 m (Figure 1a). Its general structure the typical 19th century of a monastic complex with an almost-rectangular courtyard and a church follows the typical 19th century of a monastic complex with an almost-rectangular courtyard and a (catholicon) in the center (Figure1b, topographic depiction: Figure1c) [ 17]. The evolution of the church (catholicon) in the center (Figure 1b, topographic depiction: Figure1c) [17]. The evolution of architectural phases of the catholicon until the 13th century is reﬂected in the Ch. Bouras drawing the architectural phases of the catholicon until the 13th century is reflected in the Ch. Bouras drawing (Figure1d) [18]. (Figure 1d) [18].

Figure 1.(a) A general photo of the Varnakova Monastery before the 2017 fire. Arrows and dotted Figure 1. (a) A general photo of the Varnakova Monastery before the 2017 ﬁre. Arrows and dotted lines lines indicate the cells’ section and the catholicon. (b) A general photo of the Varnakova Monastery indicate the cells’ section and the catholicon. (b) A general photo of the Varnakova Monastery after the afterﬁre – the yellow fire dotted– yellow lines dotted indicate lines the indicate burnt cells’the burnt area. cells’ (c) A area topographical.(c) A topographical depiction of depiction the complex of the by complexRural and by Survey Rural Engineer and Survey S. Zacharopoulos Engineer S. Zacharopoulos [19]. (d) The[19 ground].(d) Theplan ground of Ch. Bouras plan of depicting Ch. Bouras the th th depictingfour architectural the four phasesarchitectural of the phases catholicon of the between catholicon the 11thbetween and 13ththe 11 centuries and 13 [18 centuries]. [18].

The first church was founded in 1077 AD by Saint Arsenios Varnakovitis. During the 11th and 12th century, it was decorated with wall paintings and on-floor marble mosaics, while the extended

Heritage 2019, 2 1236

HeritageThe2019 first, 2 FOR church PEER REVIEW was founded in 1077 AD by Saint Arsenios Varnakovitis. During the 11th and4 12th century, it was decorated with wall paintings and on-floor marble mosaics, while the extended renovationrenovation of 1148–11511148–1151 waswas fosteredfostered byby thethe Byzantine Byzantine emperors emperors of of the the Comnenian Comnenian Dynasty, Dynasty, some some of ofwhom whom were were buried buried beneath beneath the monastery,the monastery, as referred as referred by A. by Orlandos A. Orlandos in his 1922in his study 1922 [ 20study]. The [20 church,]. The churchas it was, as at it the was time at ofthe Orlandos’s time of Orlandos’s inspection, inspection, is referred is to referred as a three-aisled to as a three basilica-aisled with basilica a narthex with and a narthex and a main altar [21]. a main altar [21]. The monastery was almost completely destroyed in 1826, as it served as refuge and base during The monastery was almost completely destroyed in 1826, as it served as refuge and base during the Greek War of Independence (1821–1830) against the Ottoman Empire. The reconstruction works the Greek War of Independence (1821–1830) against the Ottoman Empire. The reconstruction works commenced soon after the liberation, in 1831, according to the plans made by the architect Andreas commenced soon after the liberation, in 1831, according to the plans made by the architect Andreas Gasparis Kalandros, an army lieutenant, and were completed by three Epirot masons by 1838. This Gasparis Kalandros, an army lieutenant, and were completed by three Epirot masons by 1838. This is is also the year of the construction of the wooden temple of the catholicon. A. Orlandos uses the term also the year of the construction of the wooden temple of the catholicon. A. Orlandos uses the term “hammered limestones” to describe the chosen building material used in this construction phase “hammered limestones” to describe the chosen building material used in this construction phase [17,21]. [17,21]. Between the years 1992–2014, the monastery was extensively renovated. On 29 January 2017, Between the years 1992–2014, the monastery was extensively renovated. On 29 January 2017, a a devastating fire started due to a boiler explosion destroyed the largest part of the monastic cells’ devastating fire started due to a boiler explosion destroyed the largest part of the monastic cells’ quarter [22]. Thus today, the monastery cells are in a state of collapse and in need of major restoration quarter [22]. Thus today, the monastery cells are in a state of collapse and in need of major restoration and rebuilding (Figure2). and rebuilding (Figure2).

Figure 2. Depiction of the Monastery cells in its present state after the fire, which indicates the Figure 2. Depiction of the Monastery cells in its present state after the ﬁre, which indicates the external, external, the intermediate, and the internal masonry. The church, as indicated, is to the east. the intermediate, and the internal masonry. The church, as indicated, is to the east.

22.2..2. Visual Visual Inspection Inspection The external (Figure(Figure2 2,, Figure Figure3 a)3a) and and the the intermediate intermediate masonry masonry (Figure (Figure2, Figure2, Figure3d,e) 3d of, and the Figure cells’ 3quartere) of the belong cells’ toquarter the 1831–1838 belong to construction the 1831–1838 phase, construction referred to phase as historic, referred masonries. to as historic It should masonries be noted. Itthat, should during be the noted aforementioned that, during constructionthe aforementioned phase, the construction intermediate phase, masonry the was, intermediate in fact, the mas masonryonry was,surrounding in fact, the the masonry internal courtyardsurrounding and the directly internal facing courtyard the Church. and directly The historical facing masonriesthe Church. of The this historicalphase were masonries constructed of this using phase two mainwere lithotypes: constructed a)using a grey-beige two main limestone, lithotypes: which a) a occasionally grey-beige limappearsestone with, which black occasionally veins and b)appears sandstones with black of greenish veins and to brownish b) sandstones shade. of Ingreenish some areas, to brownish a black shade.compact In stonesome has area alsos, a beenblack sporadically compact stone used. has At also certain been areas sporadically of the intermediate used. At certain masonry, areas relatively of the intermediaterecent repointing masonry, seems relatively to have taken recent place. repointing seems to have taken place. The internal masonry (Figure 2, Figure 3c, and Figure 3f), which faces the courtyard and the church today, was added during the major reconstruction that took place after 1992. This aims to improve residential functionality. This new restoration masonry, which is referred to as new masonry, was raised on an already existing base (19th century construction phase). The building stone used for the construction of the interior masonry was a freshly-quarried grey-beige limestone, which greatly resembles the limestone used in the 19th century construction phase masonries.

Heritage2019, 2 FOR PEER REVIEW 5

The visual inspection of the fire-exposed masonries allowed for the following observations: i) The exterior façade of the external masonry (Figure 3a), which faces to the west, does not present any apparent damage from the fire. The interior plastered facade (Figure 3b) has clearly been affected, which displays blackened areas from the smoke produced during the fire, while partial plaster detachments can be noticed. Unfortunately, the total destruction of the wooden floors restricted access to these areas during this preliminary study. ii) The historic intermediate masonry seems to have been severely affected by the fire (Figure 3d and Figure 3e). Its building materials present several damages and stone decay patterns such as color change, spalling, and flaking (Figure 4a). The western façade of this masonry (interior of domestic cells) was covered in plaster including remnants that still remain. iii) Finally, the new masonry was strongly affected by the fire on its western façade (the one facing theHeritage cells2019 corridor, 2 ). However, its eastern façade (the one facing the courtyard) seems to be unaffected1237 (Figure 3f, Figure 4b).

Figure 3. Different areas of the cells quarter: (a) the exterior of the external wall (historic masonry), Figure 3. Different areas of the cells quarter: (a) the exterior of the external wall (historic masonry), (b) the plastered interior façade of the external wall presenting severe damages from the fire (historic (b) the plastered interior façade of the external wall presenting severe damages from the fire (historic masonry), (c) the lower part of the internal wall facing the courtyard (new masonry raised on a historic masonry), (c) the lower part of the internal wall facing the courtyard (new masonry raised on a historic base), (d) the historic intermediate masonry,(e) detail from the fire-affected intermediate masonry, base), (d) the historic intermediate masonry, (e) detail from the fire-affected intermediate masonry, and and (f) the new masonry of the cells’ quarter: its exterior façade seems intact after the fire, while its (f) the new masonry of the cells’ quarter: its exterior façade seems intact after the fire, while its interior interior has suffered fire damage. has suffered fire damage.

The internal masonry (Figure2, Figure3c,f), which faces the courtyard and the church today, was added during the major reconstruction that took place after 1992. This aims to improve residential functionality. This new restoration masonry, which is referred to as new masonry, was raised on an already existing base (19th century construction phase). The building stone used for the construction of the interior masonry was a freshly-quarried grey-beige limestone, which greatly resembles the limestone used in the 19th century construction phase masonries. The visual inspection of the fire-exposed masonries allowed for the following observations: i. The exterior façade of the external masonry (Figure3a), which faces to the west, does not present any apparent damage from the fire. The interior plastered facade (Figure3b) has clearly been affected, which displays blackened areas from the smoke produced during the fire, while partial plaster detachments can be noticed. Unfortunately, the total destruction of the wooden floors restricted access to these areas during this preliminary study. ii. The historic intermediate masonry seems to have been severely affected by the fire (Figure3d,e).

Its building materials present several damages and stone decay patterns such as color change, spalling, and flaking (Figure4a). The western façade of this masonry (interior of domestic cells) was covered in plaster including remnants that still remain. iii. Finally, the new masonry was strongly affected by the fire on its western façade (the one facing the cells corridor). However, its eastern façade (the one facing the courtyard) seems to be unaffected (Figure3f, Figure4b). Heritage2019, 2 FOR PEER REVIEW 5

The visual inspection of the fire-exposed masonries allowed for the following observations: i) The exterior façade of the external masonry (Figure 3a), which faces to the west, does not present any apparent damage from the fire. The interior plastered facade (Figure 3b) has clearly been affected, which displays blackened areas from the smoke produced during the fire, while partial plaster detachments can be noticed. Unfortunately, the total destruction of the wooden floors restricted access to these areas during this preliminary study. ii) The historic intermediate masonry seems to have been severely affected by the fire (Figure 3d and Figure 3e). Its building materials present several damages and stone decay patterns such as color change, spalling, and flaking (Figure 4a). The western façade of this masonry (interior of domestic cells) was covered in plaster including remnants that still remain. iii) Finally, the new masonry was strongly affected by the fire on its western façade (the one facing the cells corridor). However, its eastern façade (the one facing the courtyard) seems to be unaffected (Figure 3f, Figure 4b).

Figure 3. Different areas of the cells quarter: (a) the exterior of the external wall (historic masonry), (b) the plastered interior façade of the external wall presenting severe damages from the fire (historic masonry), (c) the lower part of the internal wall facing the courtyard (new masonry raised on a historic base), (d) the historic intermediate masonry,(e) detail from the fire-affected intermediate masonry, Heritageand2019 (f), 2the new masonry of the cells’ quarter: its exterior façade seems intact after the fire, while its 1238 interior has suffered fire damage.

Figure 4. (a) Collapse of building materials from the upper part of the intermediate masonry. (b) Discolorations and scaling of the beige-grey limestone on the interior (western) façade of the new masonry.

2.3. Investigation Techniques and Methodology A variety of techniques were used in the current study, which aim to evaluate the preservation state of the masonries and their building materials comprising them, after the effect of fire. Due to the different exposure of the building materials to the fire event, it was possible to select samples with a varying degree of fire damage, which is affected and not affected. This facilitated the evaluation of the fire effect on the examined building materials. Analytical techniques were applied in the lab after sampling and were used in order to characterize the building materials and diagnose their decay due to the fire effect. Optical microscopy (OM) was employed for the petrographic characterization of the stone samples, using a ZEISS-AXIOSKOP 40 microscope, while a ProgRes-C14PLUS camera and ProgRes Capture Pro2.1 software were used for taking and editing microphotographs of the samples. Digital Microscopy (DM, i-scope, Moritex) was applied to examine the morphology and texture of the mortars under investigation. X-Ray Diffraction (XRD), using an Advanced D8 Diffractometer by Bruker, was applied for the identification of the crystalline compounds of selected stone and mortar samples, which aim to ascertain the mineralogical transformations due to the fire. Simultaneous Differential Thermal and Thermogravimetric Analysis (DTA/TG, Regulus 2500, Netzsch) was implemented on selected stone and mortar samples for the qualitative and quantitative determination of their compounds. The analyses were carried out at 30–1000 ◦C temperature range with a heating rate of 10 ◦C/min in a nitrogen atmosphere. Selected stone and mortar samples also underwent total water immersion tests [23], which aim to estimate the porosity accessible to water by total immersion for each sample, their water absorption capacity and apparent density, and, consequently, the alterations of these characteristics due to the fire effect. In addition to the analytical techniques, non destructive techniques (NDTs) were applied in situ. In particular, Schmidt hammer rebound tests (SHR) were conducted on several building stones from various areas of the cell masonries in order to evaluate surface hardness of the various lithotypes [24–26] as well as its modification due to the fire effect. A Proceq N-type hammer (impact energy: 2.207 N m) was used, which followed the operation guidelines [27]. An average of 10 impacts was performed on each of the examined building elements. All tests were performed with the hammer held perpendicular to the examined surface. For the statistical elaboration of the results, both the mean value (Rm) and the median value (RM) were calculated, while, in order to compare the dispersion of the measurements, the coefficient of variation (COV) was also estimated. Infra-Red Thermography (IRT, B200, FLIR) was used to evaluate the preservation state of the masonry as a whole, by investigating representative areas [28–30]. The passive approach implemented aimed to reveal any temperature gradations of the different building materials consisting of the masonry after the fire effect. Heritage 2019, 2 1239

HeritageThe2019 methodological, 2 FOR PEER REVIEW approach followed in this study is schematically represented in Figure57. The evaluation was conducted on two levels: building elements scale and masonry scale. The evaluation in its current condition (fire-affected). The evaluation on building material scale assists in the on material level is necessary in order to succeed in the evaluation on the masonry level, since the interpretation of the IRT results and, consequently, in the assessment of the overall preservation state individual materials comprise and, therefore, aﬀect the overall performance of the masonry. of the masonry as a whole.

Figure 5. Diagram summarizing the methodological approach followed in this study. Figure 5. Diagram summarizing the methodological approach followed in this study.

3. ResultsThe evaluation and Discussion on building materials scale was conducted both in lab and in situ. The in lab evaluation aimed to reveal petrography and texture characteristics of the various building materials,3.1. Characterization their mineralogical of the Building and chemicalMaterials composition, and physical characteristics, such as porosity, bulk density, and water uptake. In the present study, OM, DM, XRD, TG/DTA, and total water 3.1.1. Investigation of the Building Stones immersion tests were conducted in lab for the previously mentioned purposes. The in situ evaluation on buildingThe stone materials samples scale,selected aimed for in to-lab estimate examination their surfaceare described hardness, in the using following SHR. table All the (Table above 1). measurementsThe stone samples were were applied collected on both from fire-a fireff ectedaffected and and seemingly seemingly una unaffectedffected stone areas elements of the inmasonry. order to achieve a comparison of their properties’ modification because of the fire event. In situ non destructive testing by IRT wasTable used 1. asDescription a tool in orderof the selected to evaluate stone the samples preservation from the statecells’ ofquarter. the masonry (masonry scale)Stone in its current condition (fire-affected). The evaluation on building material scale assists in the interpretationSample of the IRT results and, consequently,Sample in the Description assessment of the overall preservation state of theCode masonry as a whole. 3. Results and DiscussionSandstone of green hue. Collected from intermediate historic masonry, which is V1 seemingly unaffected by the fire area (based on visual inspection). Second most 3.1. Characterization of the Building Materialsabundant stone type used in the monument. Sandstone sample, collected from intermediate historic masonry, from a fire- V2 3.1.1. Investigation of the Building Stones affected area of the masonry. The stone samplesCompact selected b forlack in-lab stone examination collected from are a described severely fire in the-affected following area table of the (Table 1). V5 The stone samples were collected from fire affectedintermediate and seemingly masonry. una ffected areas of the masonry. Grey-beige limestone from the new masonry (internal wall), which is a seemingly V6 unaffected by the fire area. V7 Grey-beige limestone from the historic intermediate masonry, affected by the fire. V8 Grey-beige limestone from the new masonry (internal wall), affected by the fire. Limestone, with two distinct zones, one black and one grey-beige, the later similar V9 to V7, unaffected by the fire (based on visual inspection). Limestone, with two distinct zones (beige-gray and black), severely affected by V10 fire.

Petrographic Examination of the Lithotypes

Heritage 2019, 2 1240

Table 1. Description of the selected stone samples from the cells’ quarter.

Stone Sample Code Sample Description Sandstone of green hue. Collected from intermediate historic masonry, which is seemingly unaffected by the V1 fire area (based on visual inspection). Second most abundant stone type used in the monument. V2 Sandstone sample, collected from intermediate historic masonry, from a fire-affected area of the masonry. V5 Compact black stone collected from a severely fire-affected area of the intermediate masonry. V6 Grey-beige limestone from the new masonry (internal wall), which is a seemingly unaffected by the fire area. V7 Grey-beige limestone from the historic intermediate masonry, affected by the fire. V8 Grey-beige limestone from the new masonry (internal wall), affected by the fire. Limestone, with two distinct zones, one black and one grey-beige, the later similar to V7, unaffected by the V9 fire (based on visual inspection). V10 Limestone, with two distinct zones (beige-gray and black), severely affected by fire.

Petrographic Examination of the Lithotypes The petrographic examination was performed on selected samples from different areas of the historic and new masonries. Samples were collected from fire affected and non-affected masonry areas based on visual inspection, which aim to trace the heating impact on the mineralogical and micro-structuralHeritage2019 characteristics, 2 FOR PEER REVIEW of the different lithotypes. 8

One of theThe main petrographic building examination stones used was performed in the historic on selected masonries samples from of 1831 different (second areas mostof the abundant stone typehistoric used and in thenew monument) masonries. Samples is a clasticwere collected calcite from sandstone fire affected consisting and non-affected of crystalline masonry and lithic fragments.areas The based sandstone on visual sampleinspection, V1 which was aim not to a traceffected the heating by the impact fire, basedon the mineralogical on the visual and inspection (Figure6a).micro According-structural to characteristi the petrographiccs of the different examination lithotypes. (Figure 6b), the crystalline fragments include One of the main building stones used in the historic masonries of 1831 (second most abundant quartz SiOstone2, feldspars type used [albite in the (NaAlSimonument)3O is8 ),a andclastic microcline calcite sandstone (KAlSi consisting3O8)], calcite of crystalline CaCO and3, muscovite lithic folia [indicativefragments. formula: The KAl sandstone2(AlSi 3sampleO10)(F,OH) V1 was2 ],not and affected grains by the of ironfire, based and zirconiumon the visual oxidesinspection/hydroxides. The lithic fraction(Figure6a). consists According of to quartzite, the petrographic limestone, examination slate, (Figure serpentine, 6b), the crystalline and sub-volcanic fragments include rockfragments. The cementingquartz SiO material2, feldspars of [albite the clasts (NaAlSi is3O of8), and sparitic microcline calcite (KAlSi nature.3O8)], calcite V2 CaCO is a3 sandstone, muscovite folia sample of a [indicative formula: KAl2(AlSi3O10)(F,OH)2], and grains of iron and zirconium oxides/hydroxides. The fire-affectedlithic area fraction of consists the historic of quartzite, masonry limestone, (Figure slate, 6serpentinec). The, and petrographic sub-volcanic rock results fragments. depict The oxidation phenomenacementing of iron-containing material of the clasts minerals. is of sparitic This calcite is macroscopicallynature. V2 is a sandstone manifested sample of a by fire the-affected reddening of the sample.area of the historic masonry (Figure 6c). The petrographic results depict oxidation phenomena of iron-containing minerals. This is macroscopically manifested by the reddening of the sample.

Figure 6. (a) Photograph of sandstone sample V1. (b) The crystalline and lithic fragments of the Figure 6. sandstone(a) Photograph are observed of. sandstone The cementing sample material V1. is of ( ba )sparitic The crystalline calcite nature and (crossed lithic Nicols fragments). (c) of the sandstonePhotograph are observed. of sandstone The sample cementing V2. (d) materialOxidation phenomena is of a sparitic of iron calcitecompounds nature into the (crossed stone Nicols). (c) Photographmass (crossed of sandstone Nicols). sample V2. (d) Oxidation phenomena of iron compounds into the stone mass (crossed Nicols). A compact black stone detected on some areas of the historic masonry is a chert (Figure7a, sample V5). This stone type is sporadically detected in the historic masonries. It was collected from a severely fire-affected area of the intermediate masonry. The stone consists of several zones. In Figure7b, three zones are discerned downwards: a) the upper fossiliferous limestone comprising sparitic fossils (mostly foraminiferal species) bound to a micritic mass, b) the intermediate zone of large rhomboid dolomite [CaMg(CO3)2] crystals,and c) the chert zone consisting of amorphous siliceous material and scattered dolomite crystals. Likewise, in Figure 7c, there is an additional zone where intense diffusion of iron hydroxides is distinguished (dark brown area, Figure 7c). This phenomenon is evident in the fossiliferous limestone zone (Figure 7f). The chert zone is depicted in

Heritage 2019, 2 1241

A compact black stone detected on some areas of the historic masonry is a chert (Figure7a, sample V5). This stone type is sporadically detected in the historic masonries. It was collected from a severely fire-affected area of the intermediate masonry. The stone consists of several zones. In Figure7b, three zones are discerned downwards: a) the upper fossiliferous limestone comprising sparitic fossils (mostly foraminiferal species) bound to a micritic mass, b) the intermediate zone of large rhomboid dolomite [CaMg(CO3)2] crystals, and c) the chert zone consisting of amorphous siliceous material and scattered dolomite crystals. Likewise, in Figure7c, there is an additional zone where intense di ffusion of ironHeritage hydroxides2019, 2 FOR isPEER distinguished REVIEW (dark brown area, Figure7c). This phenomenon is evident in9 the fossiliferousHeritage2019 limestone, 2 FOR PEER zone REVIEW (Figure 7f). The chert zone is depicted in Figure7d–e where joints filled9 withFigure carbonate 7d–7 materiale where joints and superfilled with fine-grained carbonate ironmaterial pyrite and crystals super fine are-grained observed. iron pyrite crystals areFigure observed. 7d–7e where joints filled with carbonate material and super fine-grained iron pyrite crystals are observed.

Figure 7. (a) Photograph of sample V5. (b,c) The stone constitutes of three to four zones (crossed Figure 7. (a) Photograph of sample V5. (b,c) The stone constitutes of three to four zones (crossed Nicols). NicolsFigure). 7. ( d()a )The Phot chertograph zone of issample traversed V5. ( byb,c ) joints The stone filled constitutes with carbonate of three content, to four while zones super (crossed fine- (d) The chert zone is traversed by joints filled with carbonate content, while super fine-grained iron grainedNicols). iron(d) The pyrite chert is, zone also, is observed traversed (crossed by joints Nicols filled). ( withe) The carbonate chert zone content, consists while of amorphous super fine- e pyritesiliceousgrained is, also, ironmaterial observed pyrite and is, (crossed scattered also, observed Nicols). iron pyrite (crossed ( ) crystals, The Nicols chert (metallographic). zone (e) The consists chert microscope, of zone amorphous consists incident of siliceous amorphous light). material (f) and scatteredThesiliceous fossiliferous material iron pyrite limestone and scattered crystals, zone iron.(metallographic The pyrite orange crystals, hue is microscope, (metallographicattributed to incident the microscope, diffusion light). of incident iron (f) The hydroxides light fossiliferous). (f) limestone(parallelThe fossiliferous zone. Nicols). The orangelimestone hue zone is. attributed The orange to hue the is di attributedffusion of to iron the hydroxides diffusion of (paralleliron hydroxides Nicols). (parallel Nicols). TheThe limestone limeston usede used in in the the new new masonry masonry (internal wall, wall, 1992 1992 reconstruction) reconstruction) is a is grey a grey-beige-beige biomicriticbiomicriticThe limestone limeston limestonee (sample used (sample in V6, the V6, visually new visually masonry firefire non-anon (internal-affected,ffected, wall, Figure Figure 1992 8a). reconstruction)8a). The The matrix matrix of is the of a greythestone stone-beige is of is of micriticmicriticbiomicritic texture, texture, limestone while while traces (sample traces of of fossils V6, fossils visually and and sparitic spariticfire non foraminiferal foraminiferal-affected, Figure species species 8a). areThe are detected matrix detected of inside the inside stone the stone theis ofstone mass.mass.micritic A great A texture,great number number while of tracesof joints, joints, of filled fossilsfilled with with and secondaryspariticsecondary foraminiferal calcite, calcite, traverse traverse species the theare stone detected stone mass mass inside(Figure (Figure the 8b). stone 8b). mass. A great number of joints, filled with secondary calcite, traverse the stone mass (Figure 8b).

Figure 8. (a) Photograph of limestone sample V6. (b) Grey-beige micritic fossiliferous limestone: the Figure 8. (a) Photograph of limestone sample V6. (b) Grey-beige micritic fossiliferous limestone: the Figurematrix 8. (a of) Photographthe stone is of of micritic limestone texture, sample while V6. traces (b) of Grey-beige fossils and micriticsparitic foraminiferal fossiliferous species limestone: are the detectedmatrix of inside the stone the isstone of micritic mass. A texture, great numberwhile traces of joints, of fossils filled and with sparitic secondary foraminiferal calcite, traverse species theare matrix of the stone is of micritic texture, while traces of fossils and sparitic foraminiferal species are stonedetected mass inside (parallel the stoneNicols). mass. A great number of joints, filled with secondary calcite, traverse the detectedstone inside mass (parallel the stone Nicols). mass. A great number of joints, ﬁlled with secondary calcite, traverse the stoneThe mass new (parallel masonry Nicols). is comprised solely of the grey-beige biomicritic limestone (V6). Sample V8 was alsoThe newcollected masonry from is the comprised new masonry solely fromof the an grey area-beige highly biomicritic affected limestoneby the fire (V6). (Figure Sample 9a), V8as indicated,was also collected not only fromfrom thethe location new masonry of the sample, from an but area from highly its pinkish affected hue by as thewell fire. This (Figure is con firmed9a), as byindicated, optical microscopy,not only from as the V8 location presents of similar the sample, characteristics but from itsto pinkishsample V6hue ( Figureas well .9b) This. However, is confirmed the effectby optical of the microscopy, fire is evident as V8 through presents the similar widening characteristics and rupture to sample of the V6 joints (Figure filled 9b) with. However, secondary the effect of the fire is evident through the widening and rupture of the joints filled with secondary

Heritage 2019, 2 1242

The new masonry is comprised solely of the grey-beige biomicritic limestone (V6). Sample V8 was also collected from the new masonry from an area highly affected by the fire (Figure9a), as indicated, not only from the location of the sample, but from its pinkish hue as well. This is confirmed by optical Heritage2019, 2 FOR PEER REVIEW 10 Heritagemicroscopy,2019, 2 FOR as V8 PEER presents REVIEW similar characteristics to sample V6 (Figure9b). However, the e ffect of the10 fire is evident through the widening and rupture of the joints filled with secondary calcite, which is an calcite,, which is an effect commonly observed in limestoneslimestones whenwhen exposedexposed toto extremeextreme temperaturestemperatures effect commonly observed in limestones when exposed to extreme temperatures (Figure9c). ((Figure 9c)..

Figure 9. ((a)) PhotographPhotograph ofof firefire--effected limestone sample V8.. ((b)) Grey--beige micritic fossiliferous Figure 9. (a) Photograph of fire-effected limestone sample V8. (b) Grey-beige micritic fossiliferous limestone: the matrix of the stone is of micritic texture, while traces of fossils and sparitic foraminiferal limestone:limestone: the the matrix matrix of of the the stone stone is is of of micritic micritic texture, texture, while while traces traces of of fossils fossils and and sparitic sparitic foraminiferal foraminiferal species are detected inside the stone mass. A great number of joints, filled with secondary calcite, speciesspecies areare detecteddetected inside inside the the stone stone mass. mass. A great A great number number of joints, of joints, filled withfilled secondary with secondary calcite, traversecalcite, traverse the stone mass (parallel Nicols). (c) Joint rupture with a width range of 15–18μm (crossed traversethe stone the mass stone (parallel mass Nicols).(parallel (cNicols) Joint). rupture (c) Joint with rupture a width with range a width of 15–18 rangeµ mof (crossed15–18μm Nicols). (crossed Nicols). One of the main construction materials of the historic masonry is a grey-beige micritic fossiliferous One of the main construction materials of the historic masonry is a grey-beige micritic limestone.One of Sample the main V7 was constru collectedction from materials a severely of damaged the historic area masonry of the intermediate is a grey masonry,-beige micritic which fossiliferous limestone. Sample V7 was collected from a severely damaged area of the intermediate fossiliferouswas clearly alimestoneffected by. Sample the fire V7 (Figure was collected10a). This from damaged a severely historical damaged stone, area as of expected, the intermediate presents masonry, which was clearly affected by the fire (Figure 10a). This damaged historical stone, as masonrysimilar petrographic, which was characteristics clearly affected to the by damaged the fire (Fig limestoneure 10a). collected This damaged from the historical new masonry stone, (V8), as expected, presents similar petrographic characteristics to the damaged limestone collected from the expected,which confirms presents macroscopic similar petrographic observations. characteristics Joint ruptures to the are damaged also evident. limestone However, collected they presentfrom the a new masonry (V8), which confirms macroscopic observations. Joint ruptures are also evident. newlarger masonry width, ranging (V8), which from 35 confirmsµm to 65 macroscopicµm (Figure 10observationsb,c). . Joint ruptures are also evident. However, they present a larger width, ranging from 35μm to 65μm (Figigure 10b, Figure 10c).

Figure 10. (a) Photograph of fire-effected limestone sample V7. (b) Grey-beige micritic fossiliferous Figure 10. ((a)) Photograph of fire fire-e-effectedffected limestone sample V7.V7. ( b) Grey-beigeGrey-beige micritic fossiliferous limestone:limestone: thethe matrixmatrix matrix ofof of thethe the stonestone stone isis is ofof of micriticmicritic micritic texture,texture, texture, whilewhile while tracestraces traces ofof of fossilsfossils fossils andand and spariticsparitic sparitic foraminiferalforaminiferal foraminiferal speciesspecies areareare detected detected inside inside the the stone stone mass. mass. A great A great number number of joints, of joints, filled withfilled secondary with secondary calcite, traversecalcite, traversethetraverse stone thethe mass stonestone (crossed massmass ( Nicols).(crossed (Nicolsc) Joint).). rupture((c)) JointJoint with rupturerupture a width with range a width of 35range to 65 ofµ 35m (crossedtoto 65μm Nicols).(crossed Nicols). In both fire-affected samples of the grey-beige fossiliferous limestone (V8 from new masonry, V7 fromInIn bboth historic firefire--a masonry),ffectedffected samplessamples stylolites ofof thethe are grey detected,grey--beige while fossilif ironerous hydroxides limestone and (V8 clay from minerals new masonry, seem to V7have from intruded historic into masonry) their mass.,, stylolites In parallel, are detected, as previously while mentioned,iron hydroxides both and samples clay presentminerals a seem distinct to havepinkish intruded color. into The their visual mass. reddening In parallel, of the as stone previously is attributed mentioned to the,, bothboth oxidation samplessamples of presentpresent ferrous aa minerals distinctdistinct pinkishand the color. diffusion The ofvisual iron reddening hydroxides of duethe stone to the is high attributed temperatures to the oxidation developed of ferrous during minerals the burning and theeventthe diffusiondiffusion (Figure of9ofb, ironiron Figure hydroxideshydroxides 10b,c). Generally, duedue toto thethe the highhigh reddening temperaturestemperatures of limestones developeddeveloped due duringduring to the thethe oxidation burningburning of eventevent iron (Figure 9b, Figure 10b, and Figure 10c). Generally, the reddening of limestones due to the oxidation (Figcompoundsure 9b, Fig beginsure 10b, at250–300 and Figure◦C, 10 butc). may Generally, not be evidentthe reddening before of the lim 400estones◦C when due theto the color oxidation change ofbecomes iron compounds more intense begins [10,31 at]. 250 This–300 discoloration °C, but may serves not asbe anevident indication before of the the 400 minimum °C when temperature the color changerange reached becomes during more the intense fire in [10,[10, this31 area.].]. This discoloration serves as an indication of the minimum temperaturetemperature rangerange reachedreached duringduring thethe firefire inin thisthis area.area. Sample V9 was collected from the historic intermediate masonry, and visual inspection indicatedindicated thatthat itit waswas notnot affectedaffected byby thethe fire.fire. Macroscopically, the lithotype presents two distinct zones of different coloration (black and grey--beige, Figure11a). The matrix of the stone is of micritic texture,texture, whilewhile tracestraces ofof fossilsfossils andand spariticsparitic foraminiferalforaminiferal speciesspecies areare detecteddetected insideinside thethe stonestone mmass. A great number of joints, filled with secondary calcite, traverse the stone mass ((Figure 11b). The grey-- beige biomicritic limestone is accompanied by laminaelaminae of of a a black black--grey fossiliferous limestone

Heritage 2019, 2 1243

Sample V9 was collected from the historic intermediate masonry, and visual inspection indicated that it was not affected by the fire. Macroscopically, the lithotype presents two distinct zones of different coloration (black and grey-beige, Figure 11a). The matrix of the stone is of micritic texture, while traces of fossils and sparitic foraminiferal species are detected inside the stone mass. A great number of joints, Heritagefilled with2019, secondary2 FOR PEER REVIEW calcite, traverse the stone mass (Figure 11b). The grey-beige biomicritic limestone11 Heritageis accompanied2019, 2 FOR PEER by laminae REVIEW of a black-grey fossiliferous limestone comprising foraminiferal species11 comprisingof sparitic texture foraminifera boundl to species a micritic of sparitic mass. Oxidized texture bou grainsnd to of a ferrous micritic minerals mass. Oxidized are, also, grains detected of comprising foraminiferal species of sparitic texture bound to a micritic mass. Oxidized grains of ferrous(Figure minerals11c). are, also, detected (Figure 11c). ferrous minerals are, also, detected (Figure 11c).

Figure 11. (a) Photograph of limestone sample V9 where two zones are distinct. (b) On the left, the grey-beige micritic fossiliferous limestone zone and, on the right, the grey-black fossiliferous Figure 11. (a) Photograph of limestone sample V9 where two zones are distinct. (b) On the left, the limestoneFigure 11. zone,(a) Photograph (parallel Nicols of limestone). (c) Grains sample of V9 iron where oxides two close zones to arethe distinct. interface (b of) On the the two left, zones the grey-beige micritic fossiliferous limestone zone and, on the right, the grey-black fossiliferous (parallelgrey-beige Nicols). micritic fossiliferous limestone zone and, on the right, the grey-black fossiliferous limestone limestonezone, (parallel zone, Nicols). (parallel (c) GrainsNicols). of (c iron) Grains oxides of close iron to oxides the interface close to of the the interface two zones of (parallel the two Nicols). zones (parallel Nicols). Based on visual inspection, a fire-affected sample of the above-described lithotype from the Based on visual inspection, a fire-affected sample of the above-described lithotype from the historic intermediate masonry was, also, examined (Figure 12, Sample V10). The visual reddening of historicBased intermediate on visual masonry inspection, was, a also,fire-affected examined sample (Figure of 12 the, Sample above- V10).described The visual lithotype reddening from the of the stone is attributed to the oxidation of ferrous minerals and the diffusion of iron hydroxides due histtheoric stone intermediate is attributed masonry to the oxidation was, also, of ferrousexamined minerals (Figure and 12, theSample diffusion V10). of The iron visual hydroxides reddening due of to to the high temperatures developed during the burning event (Figures12 a, 12b, and 12c), as also the highstone temperatures is attributed developedto the oxidation during of theferrous burning minerals event and (Figure the 12diffusiona–c), as of also iron noticed hydroxides in samples due noticed in samples V8 and V7. The capillary rupture observed in the stone mass constitutes a micro- V8to the and high V7. Thetemperatures capillary rupturedeveloped observed during in the the burning stone mass event constitutes (Figures12 a micro-structurala, 12b, and 12c), change,as also structural change, as an additional result of the fire impact on the beige-grey limestone zone. The noticedas an additional in samples result V8 and of the V7 fire. The impact capillary on the rupture beige-grey observed limestone in the zone.stone mass The black-grey constitutes limestone a micro- black-grey limestone zone does not present any significant difference from the V9 sample, which was structuralzone does notchange, present as an any additional significant result difference of the from fire the impact V9 sample, on the which beige- wasgrey not limestone affected byzone the. The fire. not affected by the fire. black-grey limestone zone does not present any significant difference from the V9 sample, which was not affected by the fire.

Figure 12. (a) Photograph of fire-affected limestone sample V10. (b) The capillary rupture of some μm Figure 12. (a) Photograph of fire-affected limestone sample V10. (b) The capillary rupture of some µm width observed in the stone mass constitutes a micro-structural change. The orange hue is due to the widthFigure observed12. (a) Photograph in the stone of massfire-affected constitutes limestone a micro-structural sample V10. change.(b) The capillary The orange rupture hue isof due some to μm the diffusion of iron hydroxides and infers the visually observed reddening of the stone (parallel Nicols). widthdiffusion observed of iron in hydroxides the stone mass and infers constitutes the visually a micro observed-structural reddening change. The of the orange stone hue (parallel is due Nicols). to the (c) Oxidation of iron compounds due to the high temperature during the fire (parallel Nicols). diffusion(c) Oxidation of iron of ironhydroxides compounds and infers due to the the visually high temperature observed reddening during the of fire the (parallel stone (parallel Nicols). Nicols). (c) Oxidation of iron compounds due to the high temperature during the fire (parallel Nicols). Mineralogical Examination of the lithotypes through X-rayX-ray DiDiffractionffraction MineralogicalMineralogical Examination examination examination of the was was lithotypes conducted conducted through only only on X - samples onray samples Diffraction of the of main the buildingmain building stones ofstones historic of historic and new masonries, sandstone samples V1, V2, and limestone samples V6, V7, and V8. The and newMineralogical masonries, examination sandstone samples was conducted V1, V2, and only limestone on samples samples of V6, the V7, main and building V8. The lithotypes stones of lithotypes were pulverized and studied through X-ray diffraction. The results are summarized in werehistoric pulverized and new andmasonries, studied sandstone through X-ray samples diffraction. V1, V2, Theand resultslimestone are summarizedsamples V6, V7, in Table and 2V8.. The Table 2. lithotypesRegarding were thepulverized green sandstone, and studied both through unaffected X-ray and diffraction. affected samples The results (V1 and are V2, summarized respectively) in Regarding the green sandstone, both unaffected and affected samples (V1 and V2, respectively) Tablepresent 2. similar mineralogical composition, with quartz as the principal mineralogical component and present similar mineralogical composition, with quartz as the principal mineralogical component and calcite,Regarding albite, muscovite, the green and sandstone, chlorite both as accessory unaffected mineralogical and affected components. samples (V1 However, and V2, inrespectively) sample V1, calcite, albite, muscovite, and chlorite as accessory mineralogical components. However, in sample present similar mineralogical composition, with quartz as the principal mineralogical component and V1, microcline is detected, while, in sample V2, it is not detected. This is in accordance to other studies calcite, albite, muscovite, and chlorite as accessory mineralogical components. However, in sample that have demonstrated that microcline does decrease as the temperature of the stone increases [32]. V1, microcline is detected, while, in sample V2, it is not detected. This is in accordance to other studies Regarding the beige-grey limestone, the only lithotype used in the new masonry and the most that have demonstrated that microcline does decrease as the temperature of the stone increases [32]. dominant in the historic ones, no alterations are noticed in relation to mineralogical composition Regarding the beige-grey limestone, the only lithotype used in the new masonry and the most between the unaffected (V6) and the affected by the fire samples, from both the new (V8) and the dominant in the historic ones, no alterations are noticed in relation to mineralogical composition historic masonry (V7). Thus, the beige-gray limestone presents calcite as its principal mineralogical between the unaffected (V6) and the affected by the fire samples, from both the new (V8) and the component and quartz as an accessory.

Heritage 2019, 2 1244 microcline is detected, while, in sample V2, it is not detected. This is in accordance to other studies that have demonstrated that microcline does decrease as the temperature of the stone increases [32].

Table 2. XRD analysis of stone samples.

Sample Code Mineralogical Composition (XRD Analysis) Principal mineralogical components Accessory mineralogical components V1 Quartz calcite, albite, microcline, muscovite, chlorite V2 Quartz calcite, albite, muscovite, and chlorite V6 Calcite quartz V7 Calcite quartz V8 Calcite quartz

Regarding the beige-grey limestone, the only lithotype used in the new masonry and the most dominant in the historic ones, no alterations are noticed in relation to mineralogical composition between the unaffected (V6) and the affected by the fire samples, from both the new (V8) and the historic masonry (V7). Thus, the beige-gray limestone presents calcite as its principal mineralogical component and quartz as an accessory.

Compositional analysis of stone samples through simultaneous thermal analysis (TG/DTA) The previously mentioned stone samples (V1, V2, V6-V8) were also examined through simultaneous thermal analysis, which aim to reveal modifications in their chemical composition due to the fire. The results are presented in Table3, which report the mass loss (%) measured in specific temperature ranges.

Table 3. TG results of stone samples.

Mass Loss (%) per Temperature Range (◦C) Stone Sample Code <120 ◦C 120–200 ◦C 200–600 ◦C >600 ◦C V1 0.20 0.04 1.19 5.44 V2 1.11 0.13 1.04 1.64 V6 0.14 0.03 0.23 39.30 V7 0.17 0.02 0.36 38.99 V8 0.12 0.01 0.16 38.92

The unaffected sandstone sample V1 presents distinct differences in relation to the affected ones by the fire sandstone sample V2. In particular, sample V2 presents a higher amount of physically bound water in comparison to V1 (increased mass loss in temperature range 30–120 ◦C), which indicates a modification in the microstructure that increases water uptake from the environment after the effect of the fire. In addition, the mass loss above 600 ◦C, which is connected to the decomposition of carbonate compounds, is also distinctly different. The affected sample V2 presents a much lower carbonate content indicating that the calcite binding material of the matrix of the sandstone was reduced due to the fire, which confirms OM observations. It should be noted that, in the DTA curve, the distinct endothermic peak connected to quartz transition from α β is noticed at ~575 C. Although the → ◦ carbonate compounds’ decomposition endothermic peak is clear in the unaffected sample (~770 ◦C), it is not discernable in the fire-affected sample V2. The beige-grey limestone of the historic and new masonries (V6-V8) seems to be unaffected by the fire in relation to its chemical composition. All samples present similar mass losses in all temperature ranges, which is independent of the fire effect. This confirms XRD results as well. The decomposition temperature of the carbonate compounds was detected above 800 ◦C.

Total Immersion Results on Stone Lithotypes Total immersion was conducted on the stone samples, which aim to evaluate their porosity (%) accessible to water through total immersion. Water Absorption Capacity (WAC%) was calculated, Heritage 2019, 2 1245 while the apparent density of the samples was calculated as well. The results are summarized in Table4.

Table 4. Total water immersion tests results – stones.

Stone Sample Code Porosity (%) Bulk Density (g/cm3) W.A.C. (%) V1 3.7 2.4 1.5 V2 6.1 2.2 2.8 V6 2.6 2.3 1.1 V7 2.9 2.1 1.4 V8 2.1 2.3 0.9

The characteristics of the unaffected and fire-affected sandstone samples display diverse values, which confirm a visual inspection observation regarding their state. In particular, the affected sample, V2, presented a higher porosity (~6% in relation to 3.7%), which indicates the creation of pores and micro-cracks. This alteration is consistent with the different mass loss values demonstrated by TG 3 measurements in temperatures above 600 ◦C. Bulk density is slightly decreased from 2.4 g/cm to 2.2 g/cm3 due to the porosity increase and transformation, which take place on a mineralogical and chemical level. The water absorption capacity of the sandstone increases after the fire-event, as expected. The later statement is also evident through the higher mass loss detected for the affected stone through TG attributed to physically bound water (mass loss <120 ◦C). The beige-grey limestone, on the other hand, seems not to be affected in relation to its physical characteristics, which is also noticed in the mineralogical and thermal analysis results. Porosity of the unaffected limestone (V6) and the fire-affected one (V8), from the new masonry, is very close (2.6% and 2.1%, respectively). Bulk density values are identical (2.3 g/cm3), while water absorption capacity of the affected and non-affected samples are also extremely close (1.1%, 0.9%). The fire-affected limestone sample from the historic intermediate masonry (V7) also presents similar values to the previously mentioned samples, as expected since they are of the same lithotype. Its slightly higher porosity (2.9%) could be attributed to ageing processes in addition to the fire effect. The same is true for the slightly lower bulk density that it exhibits (2.1 g/cm3) in relation to the new masonry fire-affected limestone V8 (2.3 g/cm3), as well as the slightly higher water absorption capacity it presents (1.4% in relation to 0.9%).

3.1.2. Investigation of the Mortars Mortar samples were collected from both the historic intermediate masonry and the new masonry. The mortar samples’ description, along with characteristic digital microscopy images depicting their surface texture, are presented in Table5. Heritage2019, 2 FOR PEER REVIEW 13

Stone Sample Porosity (%) Bulk density (g/cm3) W.A.C. (%) Code V1 3.7 2.4 1.5 V2 6.1 2.2 2.8 V6 2.6 2.3 1.1 V7 2.9 2.1 1.4 V8 2.1 2.3 0.9 The characteristics of the unaffected and fire-affected sandstone samples display diverse values, which confirm a visual inspection observation regarding their state. In particular, the affected sample, V2, presented a higher porosity (~6% in relation to 3.7%), which indicates the creation of pores and micro-cracks. This alteration is consistent with the different mass loss values demonstrated by TG measurements in temperatures above 600°C. Bulk density is slightly decreased from 2.4 g/cm3 to 2.2 g/cm3 due to the porosity increase and transformation, which take place on a mineralogical and chemical level. The water absorption capacity of the sandstone increases after the fire-event, as expected. The later statement is also evident through the higher mass loss detected for the affected stone through TG attributed to physically bound water (mass loss <120 °C). The beige-grey limestone, on the other hand, seems not to be affected in relation to its physical characteristics, which is also noticed in the mineralogical and thermal analysis results. Porosity of the unaffected limestone (V6) and the fire-affected one (V8), from the new masonry, is very close (2.6% and 2.1%, respectively). Bulk density values are identical (2.3 g/cm3), while water absorption capacity of the affected and non-affected samples are also extremely close (1.1%, 0.9%). The fire-affected limestone sample from the historic intermediate masonry (V7) also presents similar values to the previously mentioned samples, as expected since they are of the same lithotype. Its slightly higher porosity (2.9%) could be attributed to ageing processes in addition to the fire effect. The same is true for the slightly lower bulk density that it exhibits (2.1 g/cm3) in relation to the new masonry fire- affected limestone V8 (2.3g/cm3), as well as the slightly higher water absorption capacity it presents (1.4% in relation to 0.9%).

3.1.2. Investigation of the Mortars Mortar samples were collected from both the historic intermediate masonry and the new Heritage 2019, 2 masonry. The mortar samples’ description, along with characteristic digital microscopy images 1246 depicting their surface texture, are presented in Table5.

Table 5.Description of the selected mortar samples from the cells’ quarter. Table 5. Description of the selected mortar samples from the cells’ quarter. Mortar Sample Description DM Image Mortar Sample CodeSample Code Sample Description DM Image

Structural soft mortar of clay texture Heritage2019, 2 FOR PEER REVIEW 14 Structuraland light soft- mortarorange of hue clay with texture blackened and light-orangeexternal hue with surface blackened due external to the surface smoke due Vm1 Heritage2019Vm1, 2 FOR PEER REVIEW 14 to the smokedeposition deposition-upper-upper part part of of the the historical historical Heritage2019, 2 FOR PEERintermediate REVIEW masonry (above 2 m). 14 intermediate masonry (above 2m). Heritage2019, 2 FOR PEER REVIEW 14

Joint mortar of cementitious texture from a severely fire-affected area from Vm2 the middle part of the new masonry Joint mortar of cementitious texture from a severely (~1.5m). Vm2 ﬁre-aﬀected area from the middle part of the new

masonry (~1.5 m).

Structural soft mortar of clay texture

andStructural dark-orange soft mortar hue from of clay the texture upper

Vm3 andpart dark of the-orange historical hue from intermediate the upper Structural soft mortar of clay texture Vm3 part ofmasonry the historical (above intermediate 2m). and dark-orange hue from the upper StructuralStructural softmasonry mortar soft mortar of (above clay textureof 2m)clay. andtexture Vm3 Vm3dark-orange part hue fromof the the historical upper part intermediate of the historical and dark-orange hue from the upper intermediatemasonry masonry (above (above 2m) 2 m).. Vm3 part of the historical intermediate

masonry (above 2m).

Structural soft mortar of clay texture

andStructural yellow softhue mortarfrom the of upper clay texture part of Vm4 the historical intermediate masonry Structuraland soft yellow mortar ofhue clay from texture the andupper yellow part hue of Structural soft(above mortar 2m) of. clay texture Vm4 Vm4 from the upperthe historical part of the intermediate historical intermediate masonry and yellow hue from the upper part of Structuralmasonry soft(above mortar 22m) m). of. clay texture Vm4 the historical intermediate masonry and yellow hue from the upper part of (above 2m). Vm4 the historical intermediate masonry

(above 2m).

Structural soft mortar of clay texture

Structuraland softStructural yellow mortar ofsofthue clay mortarfrom texture the of andupper clay yellow texture part hue of

Vm5 Vm5 from theand upperthe yellow historical part ofhue the intermediate from historical the upper intermediate masonry part of Structuralmasonry soft (above mortar 2 m). of clay texture Heritage2019Vm5, 2 FOR PEER theREVIEW historical (above intermediate 2m). masonry 15 and yellow hue from the upper part of Structural soft(above mortar 2m) of. clay texture Heritage2019Vm5, 2 FOR PEERthe REVIEW historical intermediate masonry 15 and yellow hue from the upper part of (above 2m). Vm5 the historical intermediate masonry

(above 2m).

In-depth mortar of mixed

In-depthcementitious mortarIn of-depth mixed and mortar cementitious clay oftexture mixed and from clay a

Vm6 texture fromseverely a severely fire- ﬁre-aaffectedﬀected area area at at the the lower lower Vm6 cementitious and clay texture from a part of the historicalIn-depth intermediate mortar masonryof mixed (< 1 m). severelypart of fire the- affectedhistorical area intermediate at the lower Vm6 cementitious and clay texture from a partIn of- depththemasonry historical mortar (<1m) ofintermediate mixed. severely fire-affected area at the lower Vm6 cementitiousmasonry and clay (<1m) texture. from a part of the historical intermediate severely fire-affected area at the lower

Vm6 masonry (<1m). part of the historical intermediate

masonry (<1m). Joint mortar of cementitious texture from a severely fire -affected area with Joint mortarJoint of cementitious mortar of cementitious texture from a texture severely blackened external surface due to ﬁre-aﬀectedfrom area a with severely blackened fire-affected external surfacearea with due Vm7 Vm7 to smokesmoke deposition deposition - lower part- lower of the part historical of the blackened external surface due to intermediatehistorical masonry intermediate (same sampling masonry area (same with Vm7 smoke deposition - lower part of the samplingVm6, area<1 m)with Vm6, <1m) historical intermediate masonry (same

sampling area with Vm6, <1m)

(above 2m).

Joint mortar of cementitious texture and pinkish hue from a severely fire- Joint mortar of cementitious texture Vm9 affected area at the lower part of the and pinkish hue from a severely fire- new masonry (<1m). Vm9 affected area at the lower part of the

new masonry (<1m).

3.1.3. Mineralogical Examination of the Mortar Samples through X-Ray Diffraction

3.1.3.XRD Mineralogical results are Examinationpresented in Tableof the 6Mortar and show Samples that the through principal X-Ray mineralogical Diffraction compo nent phase of the clay textured mortars (Vm1, Vm3, Vm4, Vm5, and Vm8) is quartz. The previously mentioned XRD results are presented in Table 6 and show that the principal mineralogical component phase samples all display an adequate presence of calcite and muscovite. Chlorite is also detected in of the clay textured mortars (Vm1, Vm3, Vm4, Vm5, and Vm8) is quartz. The previously mentioned samples Vm4, Vm5, and Vm8, while microcline is detected in Vm5 and Vm8. In Vm5, the presence samples all display an adequate presence of calcite and muscovite. Chlorite is also detected in of microcline is increased. Albite is detected with a high presence in samples Vm1 and Vm5 and with samples Vm4, Vm5, and Vm8, while microcline is detected in Vm5 and Vm8. In Vm5, the presence a lower presence in sample Vm3. of microcline is increased. Albite is detected with a high presence in samples Vm1 and Vm5 and with a lower presence in sample Vm3.

Heritage2019, 2 FOR PEER REVIEW 15

Joint mortar of cementitious texture from a severely fire-affected area with Joint mortar of cementitious texture blackened external surface due to from a severely fire-affected area with Vm7 smoke deposition - lower part of the blackened external surface due to historical intermediate masonry (same Heritage 2019, 2 Vm7 smoke deposition - lower part of the 1247 sampling area with Vm6, <1m) historical intermediate masonry (same

sampling area with Vm6, <1m)

Table 5. Cont.

Mortar Sample Code Sample Description DM Image

Structural soft mortar of clay texture and yellow hue from the upper part of Structural soft mortar of clay texture Vm8 the historical intermediate masonry and yellow hue from the upper part of Structural soft mortar of(above clay texture 2m). and yellow hue Vm8 Vm8 from the upperthe historical part of the intermediate historical intermediate masonry

masonry(above 22m) m)..

Joint mortar of cementitious texture and pinkish hue from a severely fire- Joint mortarJoint of cementitious mortar of cementitious texture and pinkish texture hue Vm9 Vm9from a severelyaffected ﬁre-a areaﬀected at the area lower at the lowerpart of part the of and pinkish hue from a severely fire- the newnew masonry masonry (<1 (<1m) m). . Vm9 affected area at the lower part of the

new masonry (<1m).

3.1.3. Mineralogical Examination3.1.3. Mineralogical of Examination the Mortar of the Samples Mortar Samples through through X-Ray X-Ray Diffraction Diﬀraction 3.1.3.XRD Mineralogical results are Examination presented in Tableof the 6Mortar and show Samples that the through principal X-Ray mineralogical Diffraction compo nent phase XRD results areof presented the clay textured in Table mortars6 (Vm1,and showVm3, Vm4, that Vm5 the, and principal Vm8) is quartz mineralogical. The previously component mentioned phase XRD results are presented in Table 6 and show that the principal mineralogical component phase samples all display an adequate presence of calcite and muscovite. Chlorite is also detected in of the clay textured mortarsof the clay textured (Vm1, mortars Vm3, (Vm1, Vm4, Vm3, Vm5, Vm4, andVm5, Vm8)and Vm8) is is quartz. quartz. The The previously previously mentioned mentioned samples Vm4, Vm5, and Vm8, while microcline is detected in Vm5 and Vm8. In Vm5, the presence samples all display an adequate presence of calcite and muscovite. Chlorite is also detected in samples all display anof microcline adequate is increased presence. Albite of calciteis detected and withmuscovite. a high presence Chloritein samples Vm1 is also and Vm5 detected and with in samples samples Vm4, Vm5, and Vm8, while microcline is detected in Vm5 and Vm8. In Vm5, the presence a lower presence in sample Vm3. Vm4, Vm5, and Vm8,of while microcline microcline is increased is. Albite detected is detected in with Vm5 a high and presence Vm8. in In sample Vm5,s Vm1 the and presence Vm5 and with of microcline is increased. Albite a is low detecteder presence in with sample a Vm3. high presence in samples Vm1 and Vm5 and with a lower presence in sample Vm3.

Table 6. XRD analysis of mortars’ samples.

Sample Code Mineralogical Composition (XRD Analysis) Principal mineralogical components Accessory mineralogical components Vm1 Quartz Calcite, albite, muscovite Vm2 Calcite Quartz, albite Vm3 Quartz Calcite, muscovite, albite Vm4 Quartz Calcite, muscovite, chlorite Vm5 Quartz Microcline, albite, calcite, muscovite, chlorite, Vm6 Quartz, calcite Muscovite Vm7 Calcite, quartz albite Vm8 Quartz calcite, muscovite, microcline, chlorite Vm9 Calcite, quartz -

Samples Vm2, Vm6, Vm7, and Vm9 are fire-affected mortars of a cementitious texture. Vm2 presents calcite as a principal mineralogical component and quartz and albite as accessory minerals. Vm6, which is of a mixed clay-cementitious texture, presents quartz and calcite as the main mineralogical components, while muscovite is also present. Vm7 presents calcite and quartz as principal mineralogical components, and albite is also detected. Vm 9 from the new masonry and an area that was severely affected by the fire presents only calcite and quartz as mineralogical components, with no secondary accessory minerals. It is thus evident, that all cementitious textured mortars present a high amount of calcite in relation to other minerals. Contrary to the cementitious textured mortars, the clayey textured mortars present quartz as principal mineralogical component, except for Vm6, which presents both quartz and calcite in high abundance, thus confirming to a certain extent the DM microscopy results regarding this sample’s mixed texture (Table5).

3.1.4. Compositional Analysis of Mortar Samples through Simultaneous Thermal Analysis (TG/DTA) Thermal analysis results are presented in Table7, where the weight loss percentages per temperature range are presented. Each temperature range corresponds to diﬀerent processes: (i) the Heritage 2019, 2 1248

mass loss observed in the <120 ◦C temperature range is due to the loss of physically bound water (H2OPBW), (ii) the mass loss detected in the range 120–200 ◦C is attributed to the loss of the crystalline water of salts (e.g., gypsum), (iii) the mass loss observed in the temperature range 200–600 ◦C is attributed to chemically-bound water (H20SWB) (e.g., dehydroxylation of clay minerals, dehydration of aluminosilicates, or calcium aluminosilicate phases etc.), (iv) the mass loss detected in the range >600 ◦C is attributed to the decomposition of carbonates and consequent CO2 loss [33]. The inverse hydraulicity index is calculated as the ratio CO2/H20SWB and is widely used in order to evaluate the hydraulicity of a mortar (lower inverse hydraulicity indexes are indicative of more hydraulic mortars) [34].

Table 7. TG results of mortars’ samples.

Mass Loss (%) per Temperature Range (◦C) Mortar Sample Code CO2/H2OCBW <120 ◦C 120–200 ◦C 200–600 ◦C >600 ◦C Vm1 1.17 0.44 4.02 7.96 1.98 Vm2 0.54 0.22 1.49 20.19 13.55 Vm3 1.00 0.25 2.84 8.32 2.93 Vm4 1.17 0.45 4.32 8.41 1.95 Vm5 1.23 0.43 3.81 7.33 1.92 Vm6 1.23 0.27 3.31 14.45 4.37 Vm7 1.37 0.51 2.99 20.93 7.00 Vm8 1.28 0.47 4.04 6.62 1.64 Vm9 1.50 0.62 3.11 18.57 5.97

All samples present mass losses attributed to physically bound water between 1% and 1.5%, except for the cementitious textured mortar Vm2, which presents 0.54%. They also present low mass losses in the temperature range of 120 to 200 ◦C, which indicates extremely low percentages of crystalline water bound to hygroscopic salts. Chemically bound water presents dispersion among the mortar samples ranging from 1.49% to 3.11% for the cementitious-textured mortars, and 2.84% to 4.32% for the clayey textured mortars. Sample Vm6, which is of a mixed texture, presents a chemically-bound water percentage higher than the cementitious-textured mortars and within the range of the clayey textured mortars. All cementitious-textured mortars present relatively high mass losses attributed to the loss of CO2 due to the decomposition of carbonate materials (ranging from 18.57% to 20.93%), while all clayey textured mortars present lower mass losses in the previously mentioned range (ranging from 6.62% to 8.41%). Sample Vm6, of a mixed texture, presents a CO2 mass loss value intermediate of the cementitious and clayey textured mortars. The mass loss values attributed to CO2 loss also indicate the siliceous and mixed carbonate and siliceous nature of the aggregates of all mortars investigated. However, the CO2/H2OCBW ratio is used to determine whether a mortar is of hydraulic nature. In the present study, it is used, in addition to the mass loss attributed to the decomposition of carbonates above 600 ◦C, to discriminate clay mortars from cement mortars aimed to ascertain their nature (Figure 13). It is clear that the “clayey” mortars (Vm1, Vm3, Vm4, Vm5 and Vm8) are all grouped in a very tight area, while the “cement textured” mortars are also grouped, but with a wider distribution. The mixed-textured sample, Vm6, falls between the two groups, in closer vicinity to the cement textured mortars. The thermal analysis results indicate that there are two types of mortar in the masonry, namely clay mortars and cement-based mortars. The thermal analysis results, in combination with the mineralogical analysis, indicate relatively high carbonation of the cement-based mortars, especially in the case of Vm2, which was located in the middle part of the new masonry. Heritage 2019, 2 1249 Heritage2019, 2 FOR PEER REVIEW 17

Figure 13. Correlation of CO2 and the CO2/H2OCBW ratio. The clay mortar group and cement-based Figure 13. Correlation of CO2 and the CO2/H2OCBW ratio. The clay mortar group and cement-based mortar group are indicated. mortar group are indicated. 3.1.5. Total Immersion Results on Mortar Samples 3.1.5. Total Immersion Results on Mortar Samples Total water immersion tests were conducted on selected mortar samples. Since it was not possibleTotal to water conduct immersion these measurements tests were on conducted all mortar on samples, selected due mortar to their samples. poor state Since of preservation it was not possibleand consequent to conduct dissolution these measurements in water, measurements on all mortar samples, were only due conducted to their poor on state samples of preservation Vm2, Vm9 and(cement-based consequent mortars), dissolution and in sample water, Vm4 measurements (clay mortar). were The only results conducted are summarized on samples in the Vm2, following Vm9 (cementtable (Table-based8). mortars), and sample Vm4 (clay mortar). The results are summarized in the following table (Table 8). Table 8. Total water immersion tests results – mortars. Table 8. Total water immersion tests results – mortars. Mortar Sample Code Porosity (%) Bulk Density (g/cm3) W.A.C. (%) Mortar SampleVm2 17.00 1.45 14.33 Porosity (%) Bulk Density (g/cm3) W.A.C. (%) Code Vm4 31.18 1.63 25.38 Vm9 22.00 1.34 17.15 Vm2 17.00 1.45 14.33 Vm4 The clay mortar (V4) presents31.18 the highest porosity accessible1.63 to water (31.18%),25.38 as expected for a mortarVm9 of this category. The22.00 cement mortars present relatively1.34 high porosity values17.15 of 17% and 22%The for samplesclay mortar Vm2 (V4) and presents Vm9, respectively. the highest However,porosity accessible this is distinctly to water lower (31.18%), than theas expected clay mortar. for aTheir mortar porosity of this values, category. in addition The cement to thermal mortars analysis present results, relatively confirm high a porosity high level values of carbonation, of 17% and which 22% foris probably samples dueVm2 to and the Vm9 fire, e respectivelyffect and the. consequentHowever, this microstructure is distinctly lower alteration. than the The clay water mortar. absorption Their porositycapacity values, of the clay in addition mortar isto thethermal highest analysis among results, the examined confirm a samples high level (25.38%), of carbonation, while the which cement is probablymortars presentdue to the similar fire valueseffect and (14.33% the consequent and 17.15%). microstructure Regarding bulk alteration. density, The the water cement absorption mortars capacityVm2 and o Vm9f the presentclay mortar values is verythe highest low for among cement the mortars, examined which samples is clearly (25.38%), due to theirwhile poor the cement state of 3 3 mortarpreservations present (1.45 similar g/cm andvalues 1.34 (14.33 g/cm%) becauseand 17.15% of the). Regarding fire effect. Thisbulk resultdensity, can the also cement be attributed mortars to Vm2a mixed and cement-lime Vm9 present composition values very oflow the for cement-based cement mortars, mortars which (which is clearly present due lowerto their bulk poor densities state of preservationthan cement (1.45 mortars) g/cm without3and 1.34 diminishing g/cm3) because the of destructive the fire effect role. This of the result fire can event. also Vmbe attributed 4, which to is 3 athe mixed clay cement mortar- examined,lime composition presents of athe bulk cement density-based of 1.63gmortars/cm (which. This present is an intermediate lower bulk valuedensities for thanhistoric cement mortars. mortars) without diminishing the destructive role of the fire event. Vm 4, which is the clay mortar examined, presents a bulk density of 1.63g/cm3. This is an intermediate value for historic 3.2. SHR Results mortars. AREA 1 (Figure 14) is the area left of the entrance to the Monastery and is mainly comprised of the beige-grey limestone and the sandstone. The results from the SHR measurements are presented in

Heritage2019, 2 FOR PEER REVIEW 18

3.2. SHR Results Heritage 2019, 2 1250 AREA 1 (Figure 14) is the area left of the entrance to the Monastery and is mainly comprised of the beige-grey limestone and the sandstone. The results from the SHR measurements are presented Table9. As already mentioned, this part of the masonry belongs to the historical phase of 1831-1838 in Table 9. As already mentioned, this part of the masonry belongs to the historical phase of 1831- and is seemingly unaﬀected by the ﬁre. The highest Rebound value corresponds to the black chert, 1838 and is seemingly unaffected by the fire. The highest Rebound value corresponds to the black which is found as an inclusion inside the limestone in this area. The grey-beige limestone presents chert, which is found as an inclusion inside the limestone in this area. The grey-beige limestone higher surface hardness index values than the sandstone examined. presents higher surface hardness index values than the sandstone examined.

Figure 14. AREA 1: The exterior of the external masonry, on the left of the entrance to the monastic Figure 14. AREA 1: The exterior of the external masonry, on the left of the entrance to the monastic cells quarter (historic(historic masonry, western western façade). façade). Bullets Bullets and and numbering numbering indicate indicate where where the SHR measurements were performed.

Table 9.9.SHRSHR values values from from the the AREA AREA 1 1 (see (see Figure 14).).

AAREAREA 1 TypeType of of Lithotype Lithotype RmRm RmedianRmedian COV%COV% 1.1.Beige Beige - grey- grey limestone limestone 51.04 1.9751.04 ± 1.97 50.00 50.001.25 ± 1.25 7.6 7.6 ± ± 2.2. Sandstone Sandstone 43.96 3.1443.96 ± 3.14 44.00 44.002.25 ± 2.25 7.14 7.14 ± ± 3. Black chert inclusion 56.92 4.72 58.50 5.3 8.30 3.Black chert inclusion ± 56.92 ± 4.72 ±58.50 ± 5.3 8.30 4. Beige - grey limestone 44a..Beige Black - calcitegrey limestone vein 47.58 14.20 42.75 7.75 29.84 ± ± 4b.4 Beige-greya. Black calcite limestone vein 50.08 5.2247.58 ± 14.20 49.50 42.753.50 ± 7.75 10.4229.84 ± ± 4b. Beige-grey limestone 50.08 ± 5.22 49.50 ± 3.50 10.42 The eastern façade of the intermediate historic masonry (AREA 2) was investigated after being divided per per height height in in two two areas: areas: below below 50 50 cm cm and and above above 1 1meter m (Figures (Figure 15 15aaand and 15 15b).b). This area was completely affected affected by the fire fire burning and presents the most intense deterioration patterns, such as extreme scalingscaling ofof thethe beige-grey beige-grey limestone limestone and and exfoliation exfoliation of theof the sandstones. sandstones. The The black black chert chert was was also aexaminedlso examined (Figure (Fig 15urea, sub-area15a, sub- 2).area This 2). latterThis latter type oftype stone of stone seems seems macroscopically macroscopically less aff lessected affected by the byfire, the but fire, presents but presents a SHR value a SHR decrease value decrease close to 50% close compared to 50% compared to the value to fromthe value the fire from una theffected fire unaffectedmasonry (Figure masonry 14—AREA (Figure 1, 14 sub-area—AREA 3). 1, In sub both-area zones, 3). the In beige-grey both zones limestone, the beige presents-grey limestone the lower presentsRebound the values lower compared Rebound to the values green compared sandstone, to which the green indicates sandstone, a higher which deterioration indicates level a higher due to deteriorationthe fire event level (Table due 10). to the fire event (Table 10).

Heritage2019, 2 FOR PEER REVIEW 19

The lower value for the sandstone, from all examined areas, was recorded in the intermediate masonryHeritage 2019 (Fig, 2 ure 15c, sub-area 6), in an area severely affected by the fire. This higher alteration of 1251the sandstone surface in this area may indicate a higher exposure of this part of the masonry at the fire.

FigureFigure 15. 15. AREAAREA 2: 2: Intermediat Intermediatee masonry masonry of of the the cells cells quarter quarter ( (historichistoric wall, wall, eastern façade). Bullets Bullets andand numbering numbering indicate indicate the the areas areas where where the the SHR SHR measurements measurements were were performed: performed: ( (aa)) Three Three sub sub-areas-areas examinedexamined below below 50 50 cm cm.. (b (b) ) Two Two sub sub-areas-areas examined examined above above 1 1 m m.. (c (c) )The The examined examined fire ﬁre affect aﬀecteded sandstonesandstone.. Bullets Bullets and and numbering numbering indicate indicate where where the the SHR SHR measurements measurements were were performed. performed. Table 10. SHR values from the AREA 2 (see Figure 15). Table 10.SHR values from the AREA 2 (see Figure 15). AREA 2- Under 50 cm Height AREA 2- Under 50 cm Height Type of Lithotype Rm Rmedian COV% Type of Lithotype Rm Rmedian COV% 1. Sandstone 42.14 2.67 42.00 3.00 6.34 1. Sandstone ±42.14 ± 2.67 ±42.00 ± 3.00 6.34 2. Black chert 30.00 3.96 30.50 4.50 13.19 ± ± 3. Beige-grey2.Black chert limestone 15.79 30.001.95 ± 3.96 16.25 30.502.00 ± 4.50 12.38 13.19 3. Beige-grey limestone ±15.79 ± 1.95 ±16.25 ± 2.00 12.38 AREA 2 –above 1 m height AREA 2 –above 1m height 4. Beige-grey limestone 17.14 1.52 17.00 1.10 8.87 4. Beige-grey limestone ±17.14 ± 1.52 ±17.00 ± 1.10 8.87 5. Sandstone 36.87 2.54 36.90 2.5 6.88 5. Sandstone ±36.87 ± 2.54 ±36.90 ± 2.5 6.88 SubSub-Area-Area 6 6. Sandstone 25.83 5.29 28.00 6.00 20.46 6. Sandstone ±25.83 ± 5.29 ±28.00 ± 6.00 20.46

TheThe lowerplastered value south for- thewestern sandstone, façade from of the all examined intermediate areas, masonry was recorded (AREA in 3) the had intermediate also been impactedmasonry (Figureby the fire15c, (Fig sub-areaure 16 6),). inThe an three area- severelylayer plaster affected has bya total the fire. width This that higher varies alteration from 3.5 of cm the (thickersandstone layer surface in the in western this area side) may t indicateo 2 cm (thinner a higher layer exposure on the of southern this part ofside). the masonryThe SHR attests the were fire. performedThe plastered in both south-westernsides of the stone façade blocks of the (west intermediate and south masonry sides) and (AREA after 3) the had full also removal been impacted of the overlyingby the fire plaster. (Figure The16). overall The three-layer results are plaster presented has a totalin Table width 11 that. During varies the from procedure 3.5 cm (thicker, the higher layer Reboundin the western values side) were to measured 2 cm (thinner at the layer points on theof the southern stone that side). were The covered SHR tests by werethe thicker performed plaster in layerboth sideson the of western the stone stone blocks side. (west This and may south indicate sides) andan increased after the full protective removal function of the overlying of the thicker plaster. plasterThe overall layer results on the aremasonry. presented The in reduction Table 11. in During the SHR the index procedure, for the the beige higher-grey Rebound limestone values is lower were comparedmeasured atto thethe pointseastern of façade the stone of the that intermediate were covered masonry by the thicker, which plaster was un layer-plastered, on the westernsince it stoneused toside. be the This external may indicate wall of anthe increased cells’ rooms protective (see Table function 10). of the thicker plaster layer on the masonry. The reduction in the SHR index for the beige-grey limestone is lower compared to the eastern façade of the intermediate masonry, which was un-plastered, since it used to be the external wall of the cells’ rooms (see Table 10).

Heritage 2019, 2 1252 Heritage2019, 2 FOR PEER REVIEW 20

Figure 16. AREA 3: South-Western façade of the intermediate masonry. The area was examined after Figure 16. AREA 3: South-Western façade of the intermediate masonry. The area was examined the plaster’s removal. (a) South façade of the intermediate masonry. (b) Western façade of the after the plaster’s removal. (a) South façade of the intermediate masonry. (b) Western façade of the intermediateintermediate masonry. masonry. Numbering Numbering indicates indicates the the corner corner-stones-stones examined examined by SHR SHR from from south south to to west west..

Table 11.SHRSHR values values from from AREA AREA 3 3 (see (see Figure Figure 1166).).

AAREA3REA3 TypeType of ofLithotype Lithotype RmRm RmedianRmedian COV% COV% 1. S1.andstone Sandstone 47.35 1.3347.35 ± 1.3347.00 0.8547.00 ± 0.85 2.82 2.82 ± ± 2. S2.andstone Sandstone 34.57 4.0234.57 ± 4.0234.50 4.5034.50 ± 4.50 11.64 11.64 ± ± 3. Beige3. Beige-- grey grey limestone limestone 42.5 5.6242.5 ± 5.62 43.00 5.2543.00 ± 5.25 13.22 13.22 ± ± 4. Beige-grey limestone 46.16 4.69 44.50 5.30 10.17 4. Beige-grey limestone ± 46.16 ± 4.69 ± 44.50 ± 5.30 10.17 In order to further estimate the fire impact on the grey compact limestone, a wall pillar from the new masonryIn order to(northern further estimatefaçade, from the fire interior impact to onexterior the grey) was compact examined limestone, (Figure a2 walland Fig pillarure from 3f). the newThis masonry part of (northern the masonry façade, was from constructed interior to between exterior) the was years examined 1992 to (Figures 2014. The2 and 2017’s3f). fire event left intactThis partthe ofexterior the masonry façade wasof the constructed masonry between(eastern thefaçade, years the 1992 one to 2014.facing The the 2017’s courtyard) fire event while left intenseintact the fire exterior effects façadeare observed of the masonryat the interior (eastern (western façade,) thefaçade, one such facing as the discoloration courtyard), while scaling, intense and detachments.fire effects are The observed area under at the examination interior (western), depicted façade, in suchFigure as 1 discoloration,7, was divided scaling, into zones and detachments.for the SHR measurementsThe area under to examination, be carried out depicted. a) a fire in Figure-affected 17 ,zone was divided(FAZ), according into zones to for visual the SHR inspection measurements with a totalto be length carried of out. approximately a) a fire-affected 8 cm, zone where (FAZ), a burnt according front, facing to visual west inspection (interior withfaçade) a total, and length a burnt of zoneapproximately, facing north 8 cm,, are where evident a burnt. b) front, A potentially facing west fire (interior unaffected façade), zone and (FUZ) a burnt, which zone, wasfacing further north, examinedare evident. both b) in A the potentially middle and fire in una theff ectedouter zoneside of (FUZ), each stone which block was, furtherand c) an examined in-between both zone in the of ~middle10 cm length. and in theThe outer total sidewidth of eachof the stone wall block,is 51 cm, and while c) an the in-between SHR measurements zone of ~10 cmwere length. conducted The total in awidth height of of the six wall stone is 51arrays cm, while(total height the SHR around measurements 60 cm). The were results conducted are presented in a height in Table of six 12 stone. A progressivearrays (total decline height in around the Rebound 60 cm). values The results is observed are presented. The most in Table affected 12. area A progressive i.e. the FAZ decline burnt infront the presentsRebound an values approximate is observed. loss Theof SHR most hardness affected areavalue i.e. of the52% FAZ–58% burnt compared front presents to the macroscopically an approximate perceivedloss of SHR sound hardness zones. value The SHR of 52%–58% values of compared the beige-grey to the limestone macroscopically, which are perceived visually soundperceived zones. as fireThe unaffected SHR values zones of the, are beige-grey similar limestone,to the ones which received are visuallyfrom the perceived fire unaffected as fire unaAREAffected 1 (Fig zones,ure 14, are Tablesimilar 9) to. the ones received from the fire unaffected AREA 1 (Figure 14, Table9).

Heritage 2019, 2 1253

Heritage2019, 2 FOR PEER REVIEW 21

Figure 17. AREA 4: New masonry (Northern façade). The pillar is built from the beige-grey limestone. Figure 17. AREA 4: New masonry (Northern façade). The pillar is built from the beige-grey limestone. The examined areas by SHR are labeled.

Table 12.12.SHRSHR values values from from the the AREA AREA 4 4 (see (see Figure Figure 1177).).

AREAAREA 4 Type of of Fire-A Fire-ffAffectedected Zone Zone RmRm RmedianRmedian COV%COV% FireFire Affected Affected Zone (FAZ)(FAZ) FAZ burnt front 23.86 ± 1.95 23.50 ± 2.50 8.18 FAZ burnt front 23.86 1.95 23.50 2.50 8.18 ± ± FAZFAZ burnt burnt zone zone 32.57 32.571.99 ± 1.99 33.5033.502.50 ± 2.50 6.10 6.10 Zone between FAZ & FUZ ±48.93±3.67 48.50± ± 3.10 7.5 Zone between FAZ & FUZ 48.93 3.67 48.50 3.10 7.5 Fire Unaffected± Zone (FUZ) ± Fire Unaffected Zone (FUZ) FUZ middle side 50.25 ± 4.80 49.50 ± 5.50 9.5 FUZ middle side 50.25 4.80 49.50 5.50 9.5 FUZ outer side 57.14± ± 3.38 56.00± ± 4.00 5.91 FUZ outer side 57.14 3.38 56.00 4.00 5.91 ± ± 3.3. IRT Inspection 3.3. IRT Inspection IRT investigation took place on several parts of the new and the historical masonry, and some characterisIRT investigationtic results are took presented place on below. several parts of the new and the historical masonry, and some characteristicThe higher results masonry are presented part of the below. new masonry (above 1m) presents lower temperature readings comparedThe higher to the masonrylower masonry part of part, the new as well masonry as an intermediate (above 1m) presents temperature lower distribution temperature width readings of 3 °comparedC (from 23 to °C the to lower27 °C), masonry (Figure 1 part,8a–18 asd). well Regarding as an intermediate the lower masonry temperature part, distributionthe lower stone width zone of presents3 ◦C (from higher 23 ◦C temperatures to 27 ◦C), (Figure (Fig ure18a–d).s 18c Regardingand 18d, zone the lower under masonry the dotted part, line) the lowercompared stone to zone the upperpresents zone higher (Figure temperatures 18c, 18d, zone (Figure above 18 c,d,the dotted zone under line), theas well dotted as a line) tighter compared temperature to the distribut upper zoneion width(Figure of 18 2 c,d,°C (from zone 26 above °C to the 28 dotted°C), whereas line), as the well upper as a zone tighter displays temperature a temperature distribution distribution width of width 2 ◦C of(from 4 °C 26 (from◦C to 23.5 28 ◦ °CC), to whereas 27.5°C ). the upper zone displays a temperature distribution width of 4 ◦C (from 23.5 ◦ItC can to 27.5be assumed◦C). by the 2 IRT images’ co-evaluation that the new masonry presents different temperature zones height-wise. In particular, lower masonry parts display higher temperature readings compared to the upper masonry parts. This indicates a difference of the fire effect on the new masonry stone blocks, according to their height and position, even though a precise pattern is not demonstrated.

Heritage 2019, 2 1254

Heritage2019, 2 FOR PEER REVIEW 22

Figure 18. (a) Photo of the examined area 1 on the new masonry (higher masonry part). (b) Figure 18. (a) Photo of the examined area 1 on the new masonry (higher masonry part). Thermography of area 1. (c) Photo of the examined area 2 on the new masonry (lower masonry part- (b) Thermography of area 1. (c) Photo of the examined area 2 on the new masonry (lower masonry area 2 is below area 1). (d) Thermography of area 2. part-area 2 is below area 1). (d) Thermography of area 2.

IRTIt can investigation be assumed of by the the intermediate 2 IRT images’ historical co-evaluation masonry that shows the new temperature masonry variations presents di amongfferent Figure 18. (a) Photo of the examined area 1 on the new masonry (higher masonry part). (b) thetemperature different zonesstone height-wise.types examined In particular, (Figures 19a lower and masonry 19b), which parts are display attributed higher to temperature the different readings degree Thermography of area 1. (c) Photo of the examined area 2 on the new masonry (lower masonry part- ofcompared the surface to the detachment/scaling upper masonry parts. due to This the indicates fire effect. a diThefference intermediate of the fire historical effect on masonry the new masonrypresents area 2 is below area 1). (d) Thermography of area 2. astone wide blocks, temperature according distribution to their height width and of 4.5 position, °C (from even 22.5 though °C to a27 precise °C ). pattern is not demonstrated. IRT investigation of the intermediate historical masonry shows temperature variations among IRT investigation of the intermediate historical masonry shows temperature variations among the different stone types examined (Figure 19a,b), which are attributed to the different degree of the the different stone types examined (Figures 19a and 19b), which are attributed to the different degree surface detachment/scaling due to the fire effect. The intermediate historical masonry presents a wide of the surface detachment/scaling due to the fire effect. The intermediate historical masonry presents temperature distribution width of 4.5 C (from 22.5 C to 27 C). a wide temperature distribution width◦ of 4.5 °C (from◦ 22.5 °C◦ to 27 °C ).

Figure 19. (a) Photo of the examined area 3 of the intermediate historical masonry. (b) Thermography of area 3.

The application of IRT at higher historic masonry parts (above 2m) display lower temperature readings compared to the lower masonry parts (Figure 20).The wooden element in the middle of the IRT imageFigure presents19. ((aa)) Photo Photo temperatures of of the the examined examined up to area area 25.5 3 3 °Cof of the, the while intermediate intermediate the masonry historical historical parts masonry masonry. above. and( (bb)) Thermograph Thermography below it reachy up to 26.8of °Carea . The 3. temperature distribution width of the masonry, except the wooden element, is about 4.5 °C . Thus, it is demonstrated that the historic masonry holds a similar pattern with the new one, as farThe as itapplication concerns tempe of IRTrature at higher distribution historic. masonryThe upper parts masonry (above parts 2m) display display lower lower temperatures temperature whenreadings compared compared to the to the lower lower parts. masonry parts (Figure 20).The wooden element in the middle of the IRT image presents temperatures up to 25.5°C , while the masonry parts above and below it reach up to 26.8 °C . The temperature distribution width of the masonry, except the wooden element, is about 4.5 °C . Thus, it is demonstrated that the historic masonry holds a similar pattern with the new one, as far as it concerns temperature distribution. The upper masonry parts display lower temperatures when compared to the lower parts.

Heritage 2019, 2 1255

The application of IRT at higher historic masonry parts (above 2 m) display lower temperature readings compared to the lower masonry parts (Figure 20). The wooden element in the middle of the IRT image presents temperatures up to 25.5 ◦C, while the masonry parts above and below it reach up to 26.8 ◦C. The temperature distribution width of the masonry, except the wooden element, is about 4.5 C. Thus, it is demonstrated that the historic masonry holds a similar pattern with the new one, Heritage◦ 2019, 2 FOR PEER REVIEW 23 Heritageas far as2019 it, concerns2 FOR PEER temperature REVIEW distribution. The upper masonry parts display lower temperatures23 when compared to the lower parts.

Figure 20. (a) Photo of the examined area 4 of the intermediate historical masonry. (b) Thermography Figure 20. (a) Photo of the examined area 4 of the intermediate historical masonry. (b) Thermography of area 4. of area 4. Furthermore, IRT investigation of the historical intermediate masonry surfaces that are Furthermore, IRTIRT investigation investigation of the of historical the historical intermediate intermediate masonry masonry surfaces surfaces that are blackened that are blackened from the fire, records these areas with higher temperature readings (Figure 21); a result in blackenedfrom the ﬁre, from records the fire these, record areass these with higherareas with temperature higher temperature readings (Figure readings 21); ( aFigure result 21 in); accordance a result in accordance with the IRT examination of the respective areas of the new masonry. Temperature accordancewith the IRT with examination the IRT of examination the respective of areas the respective of the new areas masonry. of the Temperature new masonry distribution. Temperature width distribution width of the blackened surface is 2 °C (from 23.5 °C to 25.5 °C ), whereas the rest of the of the blackened surface is 2 ◦C (from 23.5 ◦C to 25.5 ◦C), whereas the rest of the masonry area under masonry area under investigation holds a temperature distribution width of 3 °C (from 22 °C to 25 investigation holds a temperature distribution width of 3 ◦C (from 22 ◦C to 25 ◦C). °C ).

Figure 21. ((aa)) Photo Photo of of the the examined examined area area 5 of the intermediate historical masonry masonry..( (bb)) Thermograph Thermographyy of area 5.

4. Conclusions Conclusions During this preliminary study,study,the the characterization characterization of of the the di differentfferent lithotypes lithotypes and and mortars mortars used used in inthe the diff erent different architectural architectural phases phases of Varnakova of Varnakova Monastery Monastery took place. took The petrographicplace. The petrographic examination examinationproved to be proved an indispensable to be an indispensable tool for characterization tool for characterization and the fire impact and the assessment fire impact on theassessment different onlithotypes the different of the lithotypes historic structure. of the historic The main structure. building The stones main building of the historic stone masonriess of the historic are a masonries beige-grey arefosilliferrous a beige-grey biomicritic fosilliferrous limestone, biomicritic as well limestone as a clastic, as well calcite as a sandstone,clastic calcite consisting sandstone, of crystallineconsisting of crystalline and lithic fragments. Other stones are also sporadically detected, such as a black chert and a black fossiliferous limestone, which appears as laminae of the beige-grey fosilliferrous biomicritic limestone. Regarding the main building stones, the petrographic examination of the fire-affected sandstone shows oxidation phenomena of iron-containing minerals due to the fire, which is macroscopically manifested by the reddening of the sample. While the fire-affected beige-grey limestone presents a pinkish hue macroscopically, due to the effect of the fire, the fire-effect is also evident through the

Heritage 2019, 2 1256 and lithic fragments. Other stones are also sporadically detected, such as a black chert and a black fossiliferous limestone, which appears as laminae of the beige-grey fosilliferrous biomicritic limestone. Regarding the main building stones, the petrographic examination of the fire-affected sandstone shows oxidation phenomena of iron-containing minerals due to the fire, which is macroscopically manifested by the reddening of the sample. While the fire-affected beige-grey limestone presents a pinkish hue macroscopically, due to the effect of the fire, the fire-effect is also evident through the widening and rupture of the joints filled with secondary calcite. The latter is more intense in the case of the limestone collected from the historical masonry. Both XRD and thermal analysis measurements indicated distinct alterations of the sandstone due to the effect of fire. The microcline peak detected in the unaffected sample is not discerned in the fire-affected sample, while the calcite content also decreased. Physical properties of the sandstone are also altered. The effect of fire has increased porosity, decreased bulk density, and increased water absorption capacity. SHR values conducted in situ on a large number of sandstone building elements, both affected and unaffected, showed a decrease of surface hardness, as expected, in accordance with the degree of the fire exposure. On the other hand, the limestone samples do not present distinct differences in their mineralogical, chemical, or physical properties due to the fire effect. However, the SHR evaluation reveals that, although relatively stable in the terms mentioned above, the effect of fire and the rupturing of the limestone’s joints filled with secondary calcite are quantitatively evident in the decreased SHR values and, therefore, the decreased surface hardness. Further research is necessary in order to ascertain whether the mechanical strength values and especially compressive strength has also decreased, as expected. This highlights the need for further sampling to obtain samples of adequate geometry and dimensions in order to accomplish mechanical strength measurements. Such a sampling campaign requires further authorization, as the Monastery is a listed historical monument and is under a strict sampling legislation. The mortars selected from the new masonry are classified as cement-based mortars, a fact which is anticipated in a recently constructed masonry. The thermal analysis and microstructure evaluation results demonstrate that these mortars are highly carbonated due to the fire effect. The mortars selected from the intermediate historical masonry belong to two categories: clay mortars and cement-based mortars. All structural mortars belong to the clay mortars category, while the final joint mortars belong to the cement-based category, which demonstrate that recent repointing had taken place using a cement-based mortar, prior to the fire event. Digital microscopy observations regarding the texture of the mortars proved to be a useful tool for an initial classification. IRT inspection of the masonries as a whole, demonstrates that the historic masonry presents relatively high temperature variations, due to the presence of diverse lithotypes, as well as the different degree of the fire effect. Similar fluctuation of temperatures is observed for the new masonry as well, despite the fact that it is entirely built from one lithotype (namely the grey-beige limestone) and can be attributed to the fire event. Both historical and new masonries present different temperature zones with their upper parts being cooler than the lower ones. This is an indication of the different effect of the fire per height. However, even though the fire effect seems to depend on position, height, and lithotype, no clear pattern is displayed. Blackened surfaces, in both the historical and new masonries, present higher temperatures, which indicate differences in the thermo-hygric behavior of the masonry in these areas, due to soot accumulation from the fire, as well as a possible stronger fire damaging effect. It is evident from the IRT results that the historical masonry presents wider temperature distribution width in comparison to the new masonry. However, the recorded temperature readings of the historical masonry are lower in comparison to the ones of the new masonry. These differences are attributed to the differentiated development of the fire event, as well as to the fire’s different effect on the historical masonry, due to the variety of building materials comprising it. Heritage 2019, 2 1257

The results of the present study provide information regarding the characterization of the building materials of the Varnakova cell’s quarter and diagnosis of their decay due to the ﬁre event. The techniques implemented have proven valuable for evaluating the present state of the masonries. These results can serve as a basis for further investigation, which are necessary in order to plan appropriate restoration and conservation strategies for the rehabilitation of the cells’ quarter in terms of sustainability.

Author Contributions: E.T.D. offered the methodology about materials characterization and nondestructive testing, and supervised the in situ investigation and sampling, all related measurements, and the overall writing of the paper. M.A. participated in the in situ investigation and sampling, as well as implemented and supervised the investigation related to thermal analysis measurements. Elaborated total water immersion tests and conducted writing and editing of the paper. I.N. conducted the writing of the original draft, elaborated the SHR measurements, and performed the DM measurements. M.T. assisted in the in situ investigation and sampling and implemented the investigation of the building materials. V.K. implemented the investigation of the building materials and performed the total water immersion tests. C.P. and G.E. implemented and supervised the microscopy and mineralogical measurements. A.M. was responsible for the scientific administration and scientific supervision of the research team, the scientific conceptualization and visualization of the project and research conducted. Funding: This research received no external funding. Acknowledgments: The authors would like to thank the Abbess and the Sisterhood of the Varnakova Monastery, as well as Ch. Alexopoulou, Civil Engineer, and S. Fidani, for inviting Professor Moropoulou and the NTUA team to undertake the diagnostic study of the Church and the Cells. In addition, the authors would like to acknowledge the Central Greece Regional unit, the Governor of Central Greece, K. Bakoyannis, the Regional Councilor, P. Aravantinou, as well as the General Directorate of Antiquities and Cultural Heritage of the Ministry of Culture and Sports of the Hellenic Republic and the Ephorate of Antiquities of Phocis, A. Psalti, for their fruitful cooperation. Lastly, the authors would like to thank NTUA Associate Professor Ch. Mouzakis for the interdisciplinary cooperation, which we anticipate to continue through a Program Agreement with the Ministry of Culture and Sports and the Central Greece Region. This research was orally presented in its initial form at the 1st International Conference TMM_CH: Transdisciplinary Multispectral Modeling and Cooperation for the Preservation of Cultural Heritage, held on 10–13 October 2018 at the Eugenides Foundation Athens, Greece under the title “The effect of fire on building materials: the case-study of the Varnakova monastery cells in Central Greece” by I. Ntoutsi, E. T. Delegou, M. Thoma, M. Apostolopoulou, Ch. Papatrechas, G. Economou, and A. Moropoulou. It has since then been modified and elaborated with further results and interpretations. Lastly, we would like to thank the anonymous reviewers for their valuable suggestions and comments during the manuscript’s revision process. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Brotóns, V.; Tomás, R.; Ivorra, S.; Alarcón, J.C. Temperature influence on the physical and mechanical properties of a porous rock: San Julian’s calcarenite. Eng. Geol. 2013, 167, 117–127. [CrossRef] 2. Kompaníková, Z.; Gomez-Heras, M.; Michˇnová, J.; Durmeková, T.; Vlˇcko,J. Sandstone alterations triggered by fire-related temperatures. Environ. Earth Sci. 2014, 72, 2569–2581. [CrossRef] 3. Ozguven, A.; Ozcelik, Y. Effects of high temperature on physico-mechanical properties of Turkish natural building stones. Eng. Geol. 2014, 183, 127–136. [CrossRef] 4. Vázquez, P.; Shushakova, V.; Gómez-Heras, M. Influence of mineralogy on granite decay induced by temperature increase: Experimental observations and stress simulation. Eng. Geol. 2015, 189, 58–67. [CrossRef] 5. Sanjurjo-Sánchez, J.; Gomez-Heras, M.; Fort, R.; Alvarez de Buergo, M.; Izquierdo Benito, R.; Angel Brud, M. Dating fires and estimating the temperature attained on stone surfaces. The case of Ciudad de Vascos (Spain). Microchem. J. 2016, 127, 247–255. [CrossRef] 6. Calia, A.; Colangiuli, D.; Lettieri, M.; Quarta, G.; Masieri, M. Microscopic techniques and a multi-analytical approach to study the fire damage of the painted stuccoes from the Petruzzelli Theatre (Bari, Southern Italy). Microchem. J. 2016, 126, 42–53. [CrossRef] 7. Freire-Lista, D.M.; Fort, R.; Varas-Muriel, M.J. Thermal stress-induced microcracking in building granite. Eng. Geol. 2016, 206, 83–93. [CrossRef] Heritage 2019, 2 1258

8. Peng, J.; Rong, G.; Cai, M.; Yao, M.; Zhou, C. Comparison of mechanical properties of undamaged and thermal-damaged coarse marbles under triaxial compression. Int. J. Rock Mech. Min. Sci. 2016, 83, 135–139. [CrossRef] 9. Vazquez, P.; Acuñab, M.; Benavente, D.; Gibeaux, S.; Navarro, I.; Gomez-Heras, M. Evolution of surface properties of ornamental granitoids exposed to high temperatures. Constr. Build. Mater. 2016, 104, 263–275. [CrossRef] 10. Ingham, J.P. Application of petrographic examination techniques to the assessment of fire-damaged concrete and masonry structures. Mater. Character. 2009, 60, 700–709. [CrossRef] 11. Martinho, E.; Mendes, M.; Dionísio, A. 3D imaging of P-waves velocity as a tool for evaluation of heat induced limestone decay. Constr. Build. Mater. 2017, 135, 119–128. [CrossRef] 12. McCabe, S.; Smith, B.J.; Warke, P.A. Exploitation of inherited weakness in fire-damaged building sandstone: The ‘fatiguing’ of ‘shocked’ stone. Eng. Geol. 2010, 115, 217–225. [CrossRef] 13. Murru, A.; Freire-Lista, D.M.; Fort, R.; Varas-Muriel, M.J.; Meloni, P. Evaluation of post-thermal shock effects in Carrara marble and Santa Caterina di Pittinuri limestone. Constr. Build. Mater. 2018, 186, 1200–1211. [CrossRef] 14. Haaland, M.M.; Friesem, D.E.; Miller, C.E.; Henshilwood, C.S. Heat-induced alteration of glauconitic minerals in the Middle Stone Age levels of Blombos Cave, South Africa: Implications for evaluating site structure and burning events. J. Arch. Sci. 2017, 86, 81–100. [CrossRef] 15. Felicetti, R. The drilling resistance test for the assessment of fire damaged concrete. Cem. Concr. Compos. 2006, 28, 321–329. [CrossRef] 16. Khoury, G.A. Effect of fire on concrete and concrete structures. Prog. Struct. Eng. Mater. 2000, 2, 429–447. [CrossRef] 17. Kardamitsi-Adami, M. The Catholikon of the Monastery Varnakova and the Architect Andreas Gasparis Kalandros. Archaiologia 2011, 6, 78–84. 18. Bouras, Ch.; Boura, L. Church Architecture in Greece around the 12th Century; Commercial Bank of Greece: Athens, Greece, 2002. 19. Zacharopoulos, S. Topographical Study of the Holy Monastery of Panagia Varnakova, Field Report; Greece, 2002. 20. Moropoulos, N. World Monuments Watch: Holy Monastery of Panagia Varnakova. 2019. submitted. 21. Orlandos, A. The Monastery of Virgin Mary of Varnakova; Doric Association of Athens: Athens, Greece, 1922. 22. Available online: https://www.newsit.gr/topikes-eidhseis/naypaktos-megali-fotia-tora-stin-iera-moni- varnakovas/1085140/ (accessed on 10 September 2018). 23. Borelli, E. ARC Laboratory Handbook: Conservation of Architectural Heritage, Historic Structures and Materials. 2. Porosity; International Centre for the Study of the Preservation and the Restoration of Cultural Property: Rome, Italy, 1999. 24. Demirdag, S.; Yavuz, H.; Altindag, R. The effect of sample size on Schmidt rebound hardness value of rocks. Int. J. Rock Mech. Min. Sci. 2009, 46, 725–730. [CrossRef] 25. Fort, R.; Alvarez de Buergo, M.; Perez-Monserrat, E.M. Non-destructive testing for the assessment of granite decay in heritage structures compared to quarry stone. Int. J. Rock Mech. Min. Sci. 2013, 61, 296–305. [CrossRef] 26. Matthews, J.A.; Owen, G.; Winkler, S.; Vater, A.E.; Wilson, P.; Mourne, R.W.; Hill, J.L. A rock-surface microweathering index from Schmidt hammer R-values and its preliminary application to some common rock types in southern Norway. Catena 2016, 143, 35–44. [CrossRef] 27. Proceq. Operating Instructions Beton Prüfhammer N/NR-L/LR; Proceq SA: Schweizenbach, Zurich, Switzerland, 2006. 28. Vasanelli, E.; Sileo, M.; Calia, A.; Aiello, M.A. Non-destructive techniques to assess mechanical and physical properties of soft calcarenitic stones. Procedia Chem. 2013, 8, 35–44. [CrossRef] 29. Vasanelli, E.; Calia, A.; Colangiuli, D.; Micelli, F.; Aiello, M.A. Assessing the reliability of non-destructive and moderately invasive techniques for the evaluation of uniaxial compressive strength of stone masonry units. Constr. Build. Mater. 2016, 124, 575–581. [CrossRef] 30. Alexakis, E.; Delegou, E.T.; Lampropoulos, K.C.; Apostolopoulou, M.; Ntoutsi, I.; Moropoulou, A. NDT as a monitoring tool of the works progress and the assessment of materials and rehabilitation interventions at the Holy Aedicule of the Holy Sepulchre. Constr. Build. Mater. 2018, 189, 512–526. [CrossRef] Heritage 2019, 2 1259

31. Borg, R.P.; Hajpál, M.; Torok, Á. The fire performance of limestone. In Applications of Structural Fire Engineering 2013; Czech Technical University: Prague, Czech Republic, 2013. 32. Sirdesai, N.N.; Mahanta, B.; Ranjith, P.G.; Singh, T.N. Effects of thermal treatment on physico-morphological properties of Indian fine-grained sandstone. Bull. Eng. Geol. Environ. 2019, 78, 883–897. [CrossRef] 33. Moropoulou, A.; Bakolas, A.; Anagnostopoulou, S. Composite materials in ancient structures. Cem. Concr. Compos. 2005, 27, 295–300. [CrossRef] 34. Moropoulou, A.; Polikreti, K.; Bakolas, A.; Michailidis, P. Correlation of physicochemical and mechanical properties of historical mortars and classification by multivariate statistics. Cem. Concr. Res. 2003, 33, 891–898. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).