The Neapolitan Yellow Tuff: An outstanding example of heterogeneity

Construction and Building Materials 136 (2017) 361–373

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

The Neapolitan Yellow Tuff: An outstanding example of heterogeneity A. Colella a,⇑, C. Di Benedetto a, D. Calcaterra a, P. Cappelletti a, M. D’Amore a, D. Di Martire a, S.F. Graziano a, L. Papa a, M. de Gennaro b, A. Langella c a

Università Federico II, Dipartimento di Scienze della Terra, dell’Ambiente e delle Risorse, DiSTAR, Via Mezzocannone 8, Napoli, Italy Centro di Competenza Regionale per lo Sviluppo ed il Trasferimento dell’Innovazione Applicata ai Beni Culturali e Ambientali ‘‘INNOVA”, Via Campi Flegrei 34, Pozzuoli, Napoli, Italy c Università del Sannio, Dipartimento di Scienze e Tecnologie, Via Port’Arsa, 11, Benevento, Italy b

h i g h l i g h t s  NYT is the most used building material in the historical architecture of Naples.  The study aims at providing an exhaustive petrophysical characterization of NYT.  The investigation was carried out on four most representative lithofacies.

a r t i c l e

i n f o

Article history: Received 12 May 2016 Received in revised form 11 January 2017 Accepted 13 January 2017 Available online 24 January 2017 Keywords: Neapolitan Yellow Tuff Lithofacies Petrophysical properties Physico-mechanical characterization Building stones Zeolites

a b s t r a c t The Neapolitan Yellow Tuff (hereafter NYT) is definitely the building material most used in the historical architecture of Naples (Southern Italy) since Greek times. The high heterogeneity of NYT, resulting from the concomitant occurrence of lithic fragments, pumices, crystals and glass cemented by crystalline or amorphous phases, the latter highly sensitive to acid environments, define a strong attitude to weathering. This research confirmed the inhomogeneous features of this volcanic formation, within the same deposit and sometimes within the same outcrop. Based on these premises, the investigation was carried out on four most representative lithofacies, according to their different textural and petrophysical features. Actually, mechanical features are governed by two concurring factors: a) the mineralogical composition; b) the texture of the rock. In the first case, the predominance of anhydrous phases (analcime + feldspar) over zeolites improves its mechanical strength. By contrast, if the mineralogical composition is constant (similar total wt.% zeolite content), the petrophysical parameters are strictly related to the texture of the rock. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Historical buildings of several ancient towns developed along the Tyrrhenian border of the Italian peninsula have been realized by using as building stones the volcanic tuffs linked to the numerous volcanic centers occurring in this area [1]. In particular, the Neapolitan Yellow Tuff (hereafter NYT), further than representing one of the largest volcaniclastic formations outcropping in Southern Italy, has always been recognized [2–5] as a historical material used as building stone since Roman and Greek ages (VI century B. C.). Worth to note is the huge monumental heritage scattered over Campania region and particularly in Naples, providing these areas a unique architectural fingerprint. To this latter respect, a survey ⇑ Corresponding author. E-mail address: [email protected] (A. Colella). http://dx.doi.org/10.1016/j.conbuildmat.2017.01.053 0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.

carried out in the whole ancient center of Naples, the largest in Europe (ca. 17 km2) revealed that NYT is one of the most diffused lithotype used ‘‘facciavista” (fair-face; Fig. 1a) [6]; in addition, NYT is also the basic building material used in Naples for no-load bearing plastered walls (Fig. 1b). NYT also constitutes the backbone of Naples, which had been built on tuff with tuff. It is therefore evident how crucial is the knowledge of this stone from several points of view, for example how it supports loadings and stresses exerted by the overhead buildings, how it behaves when used as building stone or how it reacts to the weathering actions in ancient (up to several or tens of centuries) monuments [7–8]. Although this formation has been thoroughly investigated from a volcanological and petrographical point of view some genetic aspects are still debated which provide uncertainties for an accurate stratigraphic description. Without going into details, one

362

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

Fig. 1. Castel dell’Ovo, XIII-XVI century, unplastered tuff (a); Palazzo Carafa di Maddaloni, XVII century, NYT emerges from plastered wall (b).

aspect seems to be common, namely, the great heterogeneity of this material, as a result of the varying chemical composition of the formation and fluctuating emplacement conditions. Actually, NYT represents a problematic volcanic deposit exhibiting a pronounced vertical and lateral variation in lithification grade, phase content, textural features, fabrics, etc. [9–10]. Studies aimed at characterizing NYT from a physical and mechanical point of view are few and, most of them, also dated [11–14]. A recent paper tried to investigate these aspects by means of a comparative study among macroporous stones [15]. Moreover, the above-mentioned high heterogeneity of the material makes difficult an exhaustive and complete record of the most important physical and mechanical parameters of NYT. The present research is an attempt at filling this gap. The wide knowledge of NYT based on a previous study [10] allowed us to have access to a very large sampling collected from well-defined stratigraphic units in different locations, representative of the entire formation. It was therefore possible to investigate and characterize the lithological units, in their different facies, trying to identify the variability range of any measured physical and mechanical parameter. This was the challenge of this study, which aims at providing an exhaustive petrophysical characterization of NYT and that can therefore represent a technical reference for this stone. The large number of physical parameters considered, some of them never reported before, and above all, the number of representative investigated samples give back reliable boundaries for any investigated parameter.

2. Geological setting The Phlaegrean Fields Volcanic District (Southern Italy) belongs to the Campania magmatic Province [16] that, further than Campi Flegrei, includes Somma–Vesuvius, Ischia, Procida, and the nearby islands of Ponza, Palmarola, Santo Stefano, Ventotene (Fig 2). Along with the Campanian Ignimbrite (CI), NYT is the most important pyroclastic product of Campi Flegrei, both for thickness of deposit and areal distribution, linked to a phreatomagmatic eruption occurred 15 Ka ybp [17–18]. The eruption was accompanied by a caldera collapse episode [19–22] and by the emplacement of a pyroclastic fall and flow sequence. The deposit can be distinguished in an unlithified member (A) and a lithified one (B) [22–23] (Fig. 3), on the basis of textural (grain size and sedimentary structures), compositional and depositional features. The Lower Member A is a succession of cineritic and pumiceous lapilli layers, locally massive or stratified with sandwave, planar and cross lamination structures; impact sags are also present. The Upper Member B, from which derive the investigated samples and described in details by Scarpati et al. [22], is a deposit made of cineritic layers with dispersed rounded pumices, locally rich in lithic fragments, interbedded with stratified, reversely graded, locally wavy layers, together with vesiculated ash beds. As far as the eruptive mechanisms are concerned, the deposits belonging to the Member A suggest the occurrence of a first phreatoplinian phase, followed by the formation of an eruptive column with the emplacement of fall deposits. The partial collapse

Fig. 2. Plio-Quaternary volcanic districts of central-southern Italy: Roman Province (Vulsini, Vico, Sabatini, Albani Hills), Ernici–Roccamonfina province, Campania Province (including Phlaegrean Fields, Mt. Somma-Vesuvius, Ponza and nearby islands) and Mount Vulture.

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

363

Fig. 3. Distribution of lithified and unlithified NYT deposits in Phlaegrean Fields and surrounding areas, modified after [24]. The inferred boundary of NYT (yellow dotted line) caldera and the presumed vent location (star) are highlighted.① ‘‘Liccarblock” quarry (Quarto); ② ‘‘Savanelli” quarry (Marano); ③ ‘‘Edificante” quarry (Chiaiano); ④ Nuovo Policlinico outcrop.

of the eruptive column would have led to the formation of pyroclastic surges. The beginning of the calderic collapse and repeated column collapses can be inferred from the deposits of the Member B which constitute the largest part of the entire erupted sequence. It occurs as a yellowish massive lithified tuff (the so-called ‘‘Tufo”) in the proximal areas, as an unlithified light grey pumiceous cinerite (‘‘Pozzolana”) in the more distal exposures [19–26] (Fig. 3). NYT composition ranges from latite (upper products from Member B) [23] to trachyte [22]. From a mineralogical point of view both members are mainly constituted by pumice, obsidian fragments, lithics and crystals of alkali feldspar (sanidine), clinopyroxene (augite), biotite and plagioclase embedded in an ash matrix. A diffuse zeolitization process affected Member B leading to the crystallization of phillipsite and, subordinately, chabazite and analcime [10].

acteristics, 3- areas are ‘‘manually” selected with Wand (tracing tool) and uniformly coloured with an automatic filling tool to highlight the entire surfaces, 4- the grayscale image is converted to binary data for successive quantitative processing). 3.2. Mineralogical investigations

3. Methods

Mineralogical composition was determined both by X-ray powder diffraction (XRPD-Panalytical X’Pert Pro; CuKa radiation, 40 kV, 40 mA, 4–80°2h scanning interval, 0.017° equivalent step size, 60 s equivalent counting time, RTMS X’Celerator detector and SEM-EDS observations (Jeol JSM 5310 equipped with an Oxford Inca X-act detector). Quantitative mineralogical analyses were performed by XRPD using the ‘‘Reference Intensity Ratio”-RIR technique [32]. Powders with grain size less than 10 lm were obtained using a McCrone micronizing mill; such particle size allows several problems to be overcome, i.e. particle statistics, primary extinction, micro absorption and preferred orientation [33].

3.1. Image analysis

3.3. Physico-mechanical investigations

Image Analysis processing is a helpful technique to obtain quantitative data about textural parameters of pyroclastic rocks [27–30]. This investigation was carried out on 40 rock slabs (10  10  2 cm) of different varieties of NYT which allowed to quantitatively evaluate the occurrence of pumices, lithic fragments and matrix within the analysed samples. In order to obtain the required accuracy of the analytical data all area measurements were performed on high-resolution digital scans (600–1200 dpi) of rock slabs using a free open source software [31]. Since a proper recognition of the pumice-to-lithic clasts boundaries were sometimes hard to achieve, an automatic procedure in the HSB colour space has been adopted (Image Analysis Workflow: 1- raw image is converted to HSB colour space, 2- saturation/brightness were adjusted for threshold the image and emphasize the textural char-

Specific gravity, expressed as kN/m3, was measured with a Hepycnometer (Micromeritics Multivolume Pycnometer 1305, ±0.1 to 0.2% accuracy) on cylindrical specimens (2.5 cm diameter; height 63 cm). Measured apparent and specific volumes allowed the open porosity to be calculated [34]. Connected porosity (a fraction of the total open porosity), pore size distribution and average pore radius were obtained with a Hgporosimeter (MIP, mercury intrusion porosimetry) (Thermo Finnigan Pascal 140, 240 and 440), at a maximum pressure of 400 MPa, able to investigate pore radii between 58 lm and 0.019 lm [35– 36]. The range of macro-mesopore refers to IUPAC [37]. Water absorption tests carried out on cubic specimens (7 cmside), following the suggestions UNI EN 13755 [38] allowed to determine the amount of water absorbed (weight%) after immer-

364

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

sion at atmospheric pressure. Specimens were saturated by progressive immersion in deionized water and then weighed at regular time intervals. The amount of capillary water absorbed, as a function of time, was measured according to UNI EN 15801 [39] on cubic shaped (5 cm-side) specimens. Ultrasonic wave velocities were recorded both on dry and wet cubic specimens (7 cm-side), according to UNI EN 14579 [40] using a BOVIAR DSP UTD 1004 Ultrasonic device, with a pair of 55 kHz transducers (diameter = 40 mm) in direct arrangement. To provide an adequate acoustic coupling between rock specimens and transducers, a thin film of hydro soluble gel (GIMA, Italy) was used. Uniaxial Compressive Strength tests (UCS) were performed on cubic specimens (7 cm-side) [41] with a Controls C5600 device, at a maximum load of 3000 kN and a constant rate load of 1 ± 0.5 MPa. Maximum load was continuously recorded by means of a data-acquisition system. Thermal expansion coefficient [42] determined on prismatic specimens (25  5  2.5 cm-side), using a mechanical device (Lonos Test S.r.l., Italy) allowed to record linear strains in 2080 °C temperature range. The influence of water on the swelling properties of the investigated geomaterials was evaluated on cubic specimens (5 cmside), following the procedures suggested by ISRM [43]; the apparatus measures a volumetric swelling strain by means of five digital micrometer sets (±1 lm resolution) arranged along the three axes (x,y,z). The prototype instrument was suitably realized by Lonos Test S.r.l. (Monza), based on a project developed by some authors of this research and granted by the CRdC INNOVA. 4. Materials and sampling As previously mentioned, NYT deposit shows large vertical and horizontal variations in terms of lithology and mineralogical composition along with a marked textural heterogeneity (pumices of variable dimensions, presence of lithic fragments, etc.). The oldest classification of NYT proposed a century ago by Dell’Erba [44] already accounts for this heterogeneity as witnessed by evocative names such as ‘‘arenoso” (literally, sandy-like), ‘‘turrunello” (nougat-like), ‘‘selvaiuolo” (from forest), ‘‘ferrigno” (ironcoloured), all indicating different macroscopic features of the deposit. This aspect definitely represented the most difficult task in order to get a representative sampling of the whole volcanic formation. For this reason, among the huge number of sites investigated in previous researches [9–10,15,45] four areas have been selected. The rationale followed for this choice was firstly based on the macroscopical features of the tuff which were then compared with the petrophysical data successively obtained. Two out of these four areas exposed those samples that could be roughly referred to as end-member facies either from a lithological or a textural point of view. The first end-member sample (hereafter MC) comes from Marano and Chiaiano quarries (north of Phlaegrean Fields; ② and ③ in Fig. 3) and was selected for its highest content of pumices; the second one (hereafter NP) comes from an abandoned quarry close to the Nuovo Policlinico (northeast of Phlaegrean Fields; ④ in Fig. 3). Within this quarry the NYT deposit is cross cut by a layer characterized by a very fine ashy matrix and by the almost total lack of pumice, lithics and scoriae. This sample represented the second end-member facies. Furthermore, two others facies, with intermediate textural features, were also collected: Q1 exhibiting approximately the same pumice-to-matrix ratio; Q2 characterized by the occurrence of millimetric accretionary lapilli. They were both collected in Quarto (‘‘Liccarblock” quarry, northwest of Phlaegrean Fields, ① in Fig. 3). These four facies (Table 1), reasonably represent the compendium of the whole NYT formation.

Table 1 Collecting area (see also Fig. 3) and abbreviation of the four investigated lithological facies. Facies

Source

Acronym

1 2 3 4

Marano and Chiaiano quarries ‘‘Liccarblock” quarry Quarto ‘‘Liccarblock” quarry Quarto Nuovo Policlinico

MC Q1 Q2 NP

From a macroscopic point of view MC facies (Fig. 4–1) is characterized by abundant angular, brownish centimeter-sized pumices and by an approximately 1:1 void-to-matrix ratio; millimetersized phenocrysts are visible in the yellow groundmass. Millimeter- to centimeter-scale dark grey lithic clasts are easily recognizable. The second facies (Q1; Fig. 4–2) contains significant amount of millimeter-sized (rarely centimeter) pumice and grey lithic clasts. Rare black phenocrysts of maximum 6 mm (likely clinopyroxenes) are easily recognizable along with minor small size crystals of feldspars. The third facies (Q2; Fig. 4–3) shows a texture characterized by fine (millimeter-sized) particle aggregates of ash-coated pumice lapilli, and accretionary lapilli, with rounded pumice cores. Accretionary lapilli are characterized by concentric ash layers; their occurrence suggests a phreatomagmatic origin [10,24]. Rare millimeter-size black phenocrysts (likely clinopyroxenes) are visible. The last facies (NP; Fig. 4–4) is characterized by a predominant fine-grained matrix. Pumice and lithics are very small in size (<1 mm) and hard to identify on macroscopic scale. 5. Results and discussion 5.1. Image analyses The image analysis performed on the above described samples allowed to confirm the remarkable macroscopical textural differences of the investigated facies. Table 2 reports the quantitative evaluation of some relevant textural features of NYT (pumice, lithic and matrix) which confirm the representativeness of the selected samples. Worth to note is the constant decrease of the pumice content from MC (40.4%) to NP (7.9%) counterbalanced by a likewise progressive increase of the matrix (MC: 43.9%; NP: 85.6%). As far as the lithics are considered their content is almost constant (about 13–15%) in the first three facies (MC, Q1 and Q2) and strongly decreases in NP (6.5%). 6. Mineralogical analyses Quantitative XRPD data (Table 3) evidence some differences, not only among the different facies, but also for samples belonging to the same facies. This is the case of MC for which, two mineralogically different assemblages have been distinguished: a first one (MCa) shows the highest content in phillipsite (53–55%) with very low amount of chabazite (3–9%); a second one (MCb) shows alkalifeldspars as prevailing phase (36%) followed by analcime and phillipsite in similar amounts (23%) and subordinate chabazite and mica (few units percent). In Q1 and Q2 phillipsite is the prevailing phase (35–37%) followed by alkali-feldspars (23–20%), chabazite (13–15%) and subordinate analcime (6–5%) and smectite (6–3%). From a mineralogical point of view NP samples are very similar to MCa with high contents of phillipsite (53%), subordinate chaba-

365

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

Fig. 4. Macroscopic images of the most representative NYT lithological facies: 1) MC; 2) Q1; 3) Q2; 4) NP.

Table 2 Percentages of pumices, lithics fragments and matrix (mean values). Samples

Pumices (%)

Lithics (%)

Matrix (%)

MC Q1 Q2 NP

40.4 22.6 14.1 7.9

15.7 13.3 12.2 6.5

43.9 64.1 73.7 85.6

zite (9%), smectite (6%), analcime (2%) and a lower amount of alkali-feldspars (14%). In all the samples and, more generally in the whole formation, the smectite content never exceeds 6%. This phase represents the first step of the weathering of volcanic glass; further chemical transformation of glass and smectite will lead to the crystallization of zeolites [45]. The zeolitization process is almost complete as in the final rock only few percent of residual glass still persist. The ‘‘amorphous” component reported in Table 3 could also include an aluminosilicate gel-like component as described by de Gennaro & Colella [46]. Authigenic phases were easily identified by SEM as shown in Fig. 5. All samples (except NP) exhibit prismatic and micrometric crystals of phillipsite (a) along with larger (up to 0.05 mm)

pseudo-cubic typically twinned rombohedral crystals of chabazite (b). Phillipsite is also evident in NP samples but only at a nanometric scale (c). Analcime is the prevalent phase in MCb (d). 7. Petrophysical properties First studies reporting petrophysical properties of NYT date back in time [44,47–48], whereas more recent and reliable researches were carried out an a wider and more representative number of samples [11,13–15,49–51] As previously stated, a large number of NYT outcrops has been sampled, scattered over the entire volcanic formation [10]. Actually, petrophysical tests were carried out on about 500 specimens. Table 4 reports a summary of all these data. Fig. 6 schematically reports the variations of the most representative petrophysical parameters, for macroporous rocks such as the considered tuff. The remarkable heterogeneity of the tuff in terms of variability of the main textural and structural features (crystals, lithics, pumices, etc.) has an inevitable impact on the dispersion of these parameters. In fact, bulk density and porosity considerably scatter in a wide range (from 8.14 to 14.65 kN/m3 and from 34.94 to 64.56%, respectively) and, as expected, are inversely correlated (Fig. 6).

Table 3 Mineralogical investigation on different NYT lithological facies(XRPD RIR method). Error in bracket.

*

Sample

Smectite

Px

Mica

Chabazite

Phillipsite

Analcime

Alkali-Feldspar

Tot

Amorph*

MCa MCb Q1 Q2 NP

4(±1) tr 6(±1) 3(±1) 6(±1)

2(±1) tr tr tr 3(±1)

tr 5(±1) tr tr 1

3(±1) 5(±1) 15(±1) 13(±2) 9(±2)

55(±3) 23(±2) 35(±1) 37(±2) 53(±3)

4(±1) 23(±2) 6(±1) 5(±1) 2(±1)

23(±3) 36(±3) 23(±3) 20(±3) 14(±3)

91(±10) 91(±9) 84(±7) 78(±9) 88(±11)

9(±7) 9(±7) 16(±10) 22(±12) 12(±3)

calculated by difference. Px = pyroxene, tr = traces.

366

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

Fig. 5. SEM micrograph of NYT: a) aggregates of prismatic phillipsite crystals (Q1); b) pseudo-cubic twinned chabazite crystal along with acicular phillipsite crystals (NP); c) nanometric phyllipsite (NP); d) analcime crystals (MCb).

Table 4 Petrophysical data of NYT formation.

Bulk density Specific gravity Compactness Capillarity absorption coefficient Imbibition capacity Open porosity Ultrasonic dry velocity Ultrasonic wet velocity Uniaxial compressive strength Linear thermal expansion coefficient Hydric dilatation

(kN/m3) (kN/m3) – (g/cm2
s1/2) (%) (%) (m/s) (m/s) (MPa) (10 6 mm/mm °C (%)

1

)

High porosity values along with the lowest bulk densities in MC are strongly depending on the high pumice content of this facies; by contrast, these parameters are just barely influenced by the mineralogical composition. Actually, MC and NP have the same total zeolite content (63%) but the highest and lowest bulk densities and porosities. As a matter of fact, as long as the matrix content increases, the porosity decreases progressively and the bulk density increases, accordingly. This aspect is even more enhanced as far as UCS is considered: NP, mostly constituted by a fine matrix (85%), accounts for the highest ever measured UCS values (30– 40 MPa). Fig. 7 well evidences the inverse relationship occurring between porosity and bulk density. All samples from MC, Q1 and Q2, as well as those from literature, are uniformly distributed along the regression line. By contrast, NP samples, although remaining on the same trend line, cluster in a peripheral area of the diagram. This aspect remarks the singularity of these samples (Table 2, Fig. 6) already attested by the parameters so far discussed, although confirming the correctness of the choice of this facies as an end-member. Considering the pumice content as a third variable of this diagram it is possible to discriminate all the samples in three categories (areas differently shaded in Fig. 7) each of them with

Specimens

Mean

Min

Max

Std. dv.

181 181 181 30 73 181 265 120 93 10 10

10.80 22.74 0.47 0.042 39.37 52.39 1760 1422 5.99 17.65 0.88

8.14 21.59

14.65 24.60

1.27 0.79

0.022 23.20 34.94 1373 1070 1.20 35.50 0.34

0.066 47,95 64.56 2617 2288 38.83 4.67 2.41

0.01 5.88 0.06 220.29 308.30 6.44 8.02 0.58

an increasing amount of pumices (Fig. 7, Table 2). The first group includes all the MC samples (pumices content >30%) characterized by porosities above 50% and bulk densities lower than 11 kN/m3. Most of literature data falls within this area. The second group includes Q1 and Q2 samples (pumices content from 10% to 30%) partially overlapping the first area (45– 55% porosity; 9,7–12 kN/m3); finally, NP samples group in a peripheral area (pumices content <10%) characterized by the lowest porosity values ( 35%) and the highest bulk densities (14–15 kN/m3). Porosity is a fundamental physical property that rules strength behavior and can have an impact on stone deterioration also depending on the manner in which the pore network allows the water penetration [52–57] or controls moisture transport mechanism [58]. Moreover, it is not possible to overlook the role of pore shape and pore size distribution that rule the frost or salt weathering susceptibility of a stone [15,59–63]. A deep investigation of the pore network for NYT by means of Hg-porosimetry (MIP) allowed to confirm a common bimodal distribution of pores [15,51] (Fig. 8 and Table 5), with the only exception of NP facies, once again showing its peculiar behavior. Due to the similarities of patterns within the single facies, Fig. 8 reports one representative curve for each of them.

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

367

Fig. 6. Boxplots of bulk density, porosity, water absorption and UCS values of NYT lithological facies (the ‘‘whiskers” are delimited by the minimum and maximum values, the box is delimited by the 25% and 75% quartile, and divided by the corresponding median value).

Fig. 7. Relationship between porosity and bulk density. Literature data of NYT refer to Evangelista & Pellegrino [50]. Linear regression: y = coefficient = 0.957.

MC and Q1 basically show the same pore size distribution with a first frequency peak well within the macropores range, and a second smaller one accounting for a distribution right above the mesomacropore domain (MC: 14% of mesopores; average pore radius

00501  +10646; R-squared

7.3 lm; Q1: 19% of mesopores; average pore radius 10.5 lm; Table 5). Quite different is the bimodal distribution of Q2 sample with a larger amount of meso-macropore (29% of mesopores; average pore

368

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

Fig. 8. Pore size distribution of the investigated lithological facies of NYT.

Table 5 Porosity parameters (Mercury intrusion porosimetry, MIP).

MC Q1 Q2 NP

n° samples

Porosity [%]

6 6 6 6

52.28 47.34 40.52 32.20

Average pore radius [lm]

Pore-radius distribution [relative volume,%] 0.001–0.01 [lm]

0.01–0.1 [lm]

0.1–1 [lm]

1–10 [lm]

10–100 [lm]

4 7 8 26

24 25 50 48

15 16 17 15

42 26 8 9

15 26 17 2

radius equal to 0.04; Table 5) and a second peak distribution clearly trending towards coarser pores. The unimodal pore sizes distribution of NP evidences the predominance of mesopores (61%) and the smallest average pore radius equal to 0.01 lm (Table 5). The pattern uniformly decreases towards macropores. A reliable explanation of this behavior appoints to the different amount of large pumice the extension of macropores and ultramacropores, although the latter are not appreciable by Hgporosimetry. NP sample almost lacks macropores whereas the latter are highly represented in MC. Water content is one of the most important factors influencing rock strength [54,64–67]. The high values of open porosity linked to the presence of a diffuse pore network favor a rapid penetration of water and confirm the high attitude of the tuff to absorb water (up to about 48%) (Fig. 6). The uniaxial compressive strength shows the widest variability among petrophysical properties (Fig. 6, Table 2), mainly due to UCS data of NP facies which provided very high values of this parameter if compared with the other three facies. As a matter of fact, UCS values scatter in a relatively narrow range from 1.2 MPa to 9.17 MPa for MC, Q1 and Q2 facies, whereas NP facies differs greatly from the previous three lithotypes, reaching values 3–4 times higher (30–40 MPa, Fig. 9). Although some point data show a clear overlapping, it is possible to identify four areas with a higher density of specific tuff lithological facies. As far as NP is considered, data points prevailingly cluster in an area characterized by higher UCS (>30 MPa)

7.33 10.51 0.04 0.01

and lower porosity values (34–37%). Q2 data points fall in a defined area with lower UCS values (4.5–9 MPa) and a narrow porosity range (45–50%). A third area (UCS values: from 2.7 to 5.5 MPa; porosity values: from 50 to 62%) hosts most Q1 samples, those from literature and MCb samples. The last area, partially overlapping the previous one, can now be distinguished from MCb samples only for their lower UCS values (MCa samples). The above considerations lead to hypothesize that the overall trend of all the investigated samples is again ruled by the texture (pumice, matrix, porosity, etc.). Although MC samples show the same textural features, they are characterized by substantially similar porosity values but by different UCS values. In this case it is reliable to sustain that the different mineralogical composition (Table 3) influences the different mechanical behavior. The above statements clearly indicate that porosity cannot be considered as the primary factor affecting the mechanical behavior of the tuff stone [68–69]. Actually, porosity is a function of a group of factors that generically are defined as rock texture (pumice, scoriae, lithics and matrix) as well as the occurrence of microporous phases [70]. It is worth to mention that 25 MPa is the conventional threshold value which separates weak rocks from hard rocks on an engineering-geological basis [71], hence highlighting the troublesome classification of NYT as a whole (Fig. 10). To this respect, NYT poses a difficulty similar to that already faced with another prominent volcaniclastic rock of Phlegrean origin, Piperno [72– 75], whose UCS values span between 5 and 67 MPa. Fig. 11 takes into account the role of the mineralogical assemblages on another parameter such as the ultrasonic velocity which

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

369

Fig. 9. Graph of UCS vs porosity. Literature data of NYT refer to Evangelista & Pellegrino [50].

Fig. 10. Comparison of different classification of rocks according to their strength (UCS values), modified after [71]. The dotted line divides weak rocks from hard rocks. The yellow highlighted area represents the whole range of UCS values for NYT.

is also directly correlated with the mechanical features of the rock. Again, a sharp distinction between MCa and MCb is evident. For similar values of bulk densities, the discriminant factor is again the mineralogical composition: lower ultrasonic velocities are associated to the zeolite rich samples (MCa) whereas the predom-

inance of analcime + alkali-feldspars over zeolites accounts for higher ultrasonic velocities (Fig. 11). By contrast, if the mineralogical composition is constant as in the case of MCa, Q2 and NP for which the overall zeolite content is around 50%, the increase of the values of petrophysical parameters, somehow linked to the

370

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

Fig. 11. Graph of p-wave velocities vs bulk density of investigated samples.

Fig. 12. Thermal shrinkage of NYT performed from 20 °C to 80 °C (left). The basic equipment includes furnace, pushrod, mechanical device with displacement transducer (LVDT), thermocouple (right).

Fig. 13. Swelling cell for unconfined dilatation test with five micrometer dial gauges mounted along three perpendicular directions (left). Graph of linear and volume dilatation percentages for NYT specimen (right).

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

mechanical features (ultrasonic velocity), is strictly related to the texture of the rock. As far as the thermal behavior of rocks is concerned, it is generally known its dependence from different factors spanning from the mineralogical composition, and their crystallographic orientations, pore type, pressure, and temperature [57,76–79]. A common behavior of the rock by heating is the dilatation. This is not true for NYT and generally for zeolitized materials which shrink during heating [15,80]. This aspect was confirmed by the thermal expansion test carried out on NYT samples which showed in the 20–80 °C a total linear contraction of about 1.4 mm/m after two heating cycles. This behavior is mainly due to the occurrence of hydrated phases (phillipsite, chabazite, subordinate analcime and smectite). Worth to note is that the shrinkage after the first cycle is not completely recovered (ca. 60%) after the cooling of the specimen. This aspect should be related to an uncomplete rehydration of the zeolites by cooling. The second cycle still evidences a further shrinkage which is almost completely recovered by cooling (Fig. 12). Fig. 13 reports a swelling common pattern with the axial strain in the three directions that ranges between 0.13 and 0.27%, typical of a rock with an isotropic texture, and corresponding to a total volume increase of 0.58%. This volume increase is related to the mineralogical composition and pore size distribution of the stone [12,15,80–87]. 8. Conclusions The Neapolitan Yellow Tuff represents the most used building stone of the historical monumental heritage of Naples although it also constitutes a significant part of the city modern architecture. Its petrophysical properties definitely have influenced the construction choices, mainly dictated by the workforce expertise developed in the different historical periods. During the second half of the last century many studies on the physico-mechanical parameters of this stone were carried out; however, a thorough and exhaustive comparison of these data with the minero-petrographical and textural features of the rock, was lacking. This analytical approach confirmed that NYT does not show homogeneous features and that, within the same deposit and sometimes within the same outcrop, different lithological facies may occur each characterized by variable features. Keeping in mind this aspects, a wide and accurate sampling [10] allowed to select the most different and representative lithological facies on which the present investigation was focused. The most remarkable aspect evidenced by this research was that the mechanical behavior of a tuff does not only depend on its physical parameters but also, and sometimes above all, on its petrophysical features. To summarize, three factors strongly affect the mechanical behavior of a tuff: pumice content, matrix content and mineralogical assemblage of the matrix. A pumice increase definitely worsens the tuff mechanical features whereas an increase of the fine matrix improves them; at a parity of content of fine matrix a further discriminant is its mineralogical composition: generally, the matrix of a tuff is the original fine glassy fraction that underwent a minerogenetic evolution towards open framework phases (for NYT, zeolites such as phillipsite and chabazite) or towards more closed framework ones (analcime or feldspars). A higher content of analcime and feldspars improves mechanical properties; by contrast, the consequent decrease of zeolites such as phillipsite and chabazite defaults and somehow compromises the main qualities of this rock such as lightness and thermal insulation, features that accounted for their massive use since ancient times. All these information are necessary to fully explain the behavior of this complex rock whenever it undergoes common stresses such

371

as variation of wet-dry or hot-cold conditions, load-bearings, etc., affecting a material mostly used for external building purposes. On this account, the knowledge of the above described rock features may represent a useful tool for any restorative action, correctly addressing the most suitable operative protocols. In case of tuff walls coated by plaster, such information may suggest the best choice of mortar mixtures in order to minimize the contrasting effects of the tuff shrinkage and plaster dilatation during the maximum exposure to the solar radiation, or the different swelling attitude of the two materials on water absorption.

Acknowledgement Work carried out using INNOVA scarl instruments licensed to DISTAR.

References [1] V. Morra, D. Calcaterra, P. Cappelletti, A. Colella, L. Fedele, R. de Gennaro, A. Langella, M. Mercurio, M. de Gennaro, Urban geology: relationships between geological setting and architectural heritage of the Neapolitan area, in: M. Beltrando, A. Peccerillo, M. Mattei, S. Conticelli, C. Doglioni (Eds.), Journal of the Virtual Explorer, vol. 36, 2010, http://dx.doi.org/10.3809/ jvirtex.2010.00261 paper 26. [2] A. Marzocchella, Le tombe eneolitiche di Napoli Materdei, Rivista di scienze preistoriche, 1980, pp. 147–163. XXXV. [3] V. Cardone, Il Tufo Nudo Nell’architettura Napoletana, CUEN, Napoli, 1990. [4] D. Calcaterra, P. Cappelletti, A. Colella, R. de Gennaro, A. Langella, M. de Gennaro, Le pietre dell’architettura storica della Campania, Arkos 4 (2) (2003) 40–46. [5] C. Colella, M. de Gennaro, R. Aiello, Use of zeolitic tuff in the building industry, in: D.L. Bish, D.W. Ming (Eds.), Rev. Mineral. Geochem., vol. 45(1), Mineralogical Society of America, Washington, 2001, pp. 551–588, http://dx. doi.org/10.2138/rmg.2001.45.16. [6] D. Calcaterra, P. Cappelletti, M. de Gennaro, A. Langella, V. Morra, I materiali lapidei del centro antico di Napoli: criteri metodologici per una cartografia dei litotipi e dei fenomeni di degrado, Geol. Appl. e Idrogeol., vol. XXX(I), 1995. Bari. [7] M. de Gennaro, M.D. Fuscaldo, Il degrado del tufo giallo napoletano. Prime osservazioni sui fenomeni di degrado dei materiali tufacei usati come pietra da costruzione, Arkos 13 (1991) 4–9. [8] M. de Gennaro, M.D. Fuscaldo, C. Colella, Weathering mechanisms of monumental tuffstone masonries in downtown Naples, Sci. Technol. Cult. Heritage 2 (1993) 53–62. [9] M. de Gennaro, P. Petrosino, M.T. Conte, R. Munno, C. Colella, Zeolite chemistry and distribution in a Neapolitan Yellow Tuff deposit, Eur. J. Mineral. 2 (6) (1990) 779–786, http://dx.doi.org/10.1127/ejm/2/6/0779. [10] M. de Gennaro, P. Cappelletti, A. Langella, A. Perrotta, C. Scarpati, Genesis of zeolites in the Neapolitan Yellow Tuff: geological, volcanological and mineralogical evidence, Contrib. Mineral. Petrol. 139 (1) (2000) 17–35, http://dx.doi.org/10.1007/s004100050571. [11] A. Pellegrino, Mechanical behavior of soft rocks under high stresses, Proc. of the 2nd Congress of the International Society of Rock Mechanics, Beograd, vol. 2, 1970, pp. 173–180. [12] A. Evangelista, Influenza del contenuto d’acqua sul comportamento del tufo giallo napoletano, Atti del XIV Convegno Nazionale di Geotecnica, Firenze, vol. 2, 1980, pp. 63–68. [13] F. Rippa, F. Vinale, Structures and mechanical behavior of a volcanic tuff, Proc. V Congress of the International Society of Rock Mechanics, ISRM, Melbourne, 1983. [14] S. Aversa, A. Evangelista, The mechanical behaviour of a pyroclastic rock: yield strength and ‘‘destructuration”effect, Rock. Mech. Rock. Eng. 31 (1) (1998) 25– 42, http://dx.doi.org/10.1007/s006030050007. [15] C. Di Benedetto, P. Cappelletti, M. Favaro, S.F. Graziano, A. Langella, D. Calcaterra, A. Colella, Porosity as key factor in the durability of two historical building stones: Neapolitan Yellow Tuff and Vicenza Stone, Eng. Geol. 193 (2015) 310–319, http://dx.doi.org/10.1016/j.enggeo.2015.05.006. [16] A. Peccerillo, Plio-Quaternary Volcanism in Italy, Petrology, Geochemistry, Geodynamics, Springer, Heidelberg, 2005, http://dx.doi.org/10.1007/3-54029092-3. [17] A.L. Deino, G. Orsi, S. de Vita, M. Piochi, The age of the Neapolitan Yellow Tuff caldera-forming eruption (Campi Flegrei caldera – Italy) assessed by 40Ar/ 39Ardating method, J. Volcanol. Geotherm. Res. 133 (2004) 157–170, http:// dx.doi.org/10.1016/S0377-0273(03)00396-2. [18] L. Fedele, D.D. Insinga, A.T. Calvert, V. Morra, A. Perrotta, C. Scarpati, 40Ar/39Ar dating of tuff vents in the Campi Flegrei caldera (southern Italy): toward a new chronostratigraphic reconstruction of the Holocene volcanic activity, Bull. Volcanol. 73 (9) (2011) 1323–1336, http://dx.doi.org/10.1007/s00445-0110478-8.

372

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373

[19] P. Di Girolamo, M.R. Ghiara, L. Lirer, R. Munno, G. Rolandi, D. Stanzione, Vulcanologia e petrologia dei Campi Flegrei, Boll. Soc. Geol. 103 (1984) 349– 413. [20] G. Orsi, M. D’Antonio, S. de Vita, G. Gallo, The Neapolitan Yellow Tuff, a largemagnitude trachytic phreatoplinian eruption: eruptive dynamics, magma withdrawal and caldera collapse, J. Volcanol. Geotherm. Res. 53 (1992) 275– 287, http://dx.doi.org/10.1016/0377-0273(92)90086-S. [21] G. Orsi, M. Piochi, L. Campajola, A. D’Onofrio, L. Gialanella, F. Terrasi, 14C geochronological constraints for the volcanic history of the Island of Ischia (Italy) over the last 5000 years, J. Volcanol. Geotherm. Res. 71 (1996) 249–257, http://dx.doi.org/10.1016/0377-0273(95)00067-4. [22] C. Scarpati, P. Cole, A. Perotta, The Neapolitan Yellow Tuff-A large volume multiphase eruption from Campi Flegrei, Southern Italy, Bull. Volcanol. 55 (5) (1993) 343–356, http://dx.doi.org/10.1007/BF00301145. [23] K. Wohletz, G. Orsi, S. de Vita, Eruptive mechanisms of the Neapolitan Yellow Tuff interpreted from stratigraphic, chemical and granulometric data, J. Volcanol. Geotherm. Res. 67 (4) (1995) 263–290, http://dx.doi.org/10.1016/ 0377-0273(95)00002-C. [24] P.D. Cole, C. Scarpati, A facies interpretation of the eruption and emplacement mechanisms of the upper part of the Neapolitan Yellow Tuff, Campi Flegrei, southern Italy, Bull. Volcanol. 55 (5) (1993) 311–326, http://dx.doi.org/ 10.1007/BF00301143. [25] M. Rosi, A. Sbrana, Phlegraean Fields, CNR - Quad. Ric. Sci., Roma, 1987, 114, pp. 169. [26] L. Valentini, B. Capaccioni, P.L. Rossi, R. Scandone, D. Sarocchi, Vent area and depositional mechanism of the Upper Member of the Neapolitan Yellow Tuff (Campi Flegrei, Italy): new insights from directional fabric through image analysis, Bull. Volcanol. 70 (9) (2008) 1087–1101, http://dx.doi.org/10.1007/ s00445-007-0191-9. [27] B. Capaccioni, L. Valentini, M.B.L. Rocchi, G. Nappi, D. Sarocchi, Image analysis and circular statistics for shape-fabric analysis: applications to lithified ignimbrites, Bull. Volcanol. 58 (7) (1997) 501–514, http://dx.doi.org/ 10.1007/s004450050158. [28] D. Karatson, O. Sztano, T. Telbisz, Preferred clast orientation in volcaniclastic mass-flow deposits; application of a new photo-statistical method, J. Sediment. Res. 72 (6) (2002) 823–835, http://dx.doi.org/10.1306/ 040402720823. [29] M.D. Higgins, Quantitative Textural Measurements in Igneous and Metamorphic Petrology, Cambridge University Press, Cambridge, UK, 2006, p. 276, http://dx.doi.org/10.1017/CBO9780511535574. [30] M. Jutzeler, A.A. Proussevitch, S.R. Allen, Grain-size distribution of volcaniclastic rocks 1: a new technique based on functional stereology, J. Volcanol. Geotherm. Res. 239–240 (2012) 1–11, http://dx.doi.org/10.1016/j. jvolgeores.2012.05.013. [31] W.S. Rasband, ImageJ, U.S. National Institutes of Health, Bethesda, Maryland, USA, 1997–2015. http://imagej.nih.gov/ij/. [32] D.L. Bish, S.J. Chipera, Accuracy in quantitative X-ray powder-diffraction analyses, Adv. X-ray Anal. 38 (1995) 47–57, http://dx.doi.org/10.1007/978-14615-1797-9_5. [33] H.P. Klug, L.E. Alexander, X-Ray Diffraction Procedures for Polycrystalline and Amorphous Materials, J. Wiley and Sons, New York, 1974. [34] ASTM D5550-14, Standard Test Method for Specific Gravity of Soil Solids by Gas Pycnometer, ASTM International, West Conshohocken, PA, 2014, http://dx. doi.org/10.1520/D5550-14. www.astm.org. [35] ASTM D4404-10, Standard Test Method for Determination of Pore Volume and Pore Volume Distribution of Soil and Rock by Mercury Intrusion Porosimetry, ASTM International, West Conshohocken, PA, 2010, http://dx.doi.org/10.1520/ D4404-10. www.astm.org. [36] International Organization for Standardization (ISO), ISO 15901-1-2005, Pore Size Distribution and Porosity of Solid Materials by Mercury Porosimetry and Gas Adsorption-Part 1: Mercury Porosimetry, ISO, Geneva, Switzerland, 2005. [37] J. Rouquerol, D. Avnir, C.W. Fairbridge, D.H. Everett, J.H. Haynes, N. Pernicone, J. D.F. Ramsay, K.S.W. Sing, K.K. Unger, Recommendations for the characterization of porous solids, Pure Appl. Chem. 66 (8) (1994) 1739– 1758, http://dx.doi.org/10.1351/pac199466081739. [38] UNI EN 13755, Natural Stone Test Methods. Determination of Water Absorption at Atmospheric Pressure, CNR-ICR, Rome, 2008 (in italian). [39] UNI EN 15801, Conservation of Cultural Property-Test MethodsDetermination of Water Absorption by Capillarity, CNR-ICR, Rome, 2010 (in italian). [40] UNI EN 14579, Natural Stone Test Methods-Determination of Sound Speed Propagation, CNR-ICR, Rome, 2005 (in italian). [41] UNI EN 1926, Natural Stone Test Methods-Determination of Uniaxial Compressive Strength, CNR-ICR, Rome, 2007 (in italian). [42] UNI EN 14581, Natural stone Test Methods-Determination of Linear Thermal Expansion Coefficient, CNR-ICR, Rome, 2005 (in italian). [43] International Society for Rock Mechanics (ISRM), Rock characterization, testing and monitoring, in: E.T. Brown (Ed.), ISRM Suggested Methods, International Society for Rock Mechanics, Pergamon Press, Oxford, 1981. [44] L. Dell’Erba, The Neapolitan Yellow Tuff. Technical-scientific study extended to quarries and landslides, in: R. Pironti (Ed.), Napoli, 1923, pp. 288. (in italian). [45] M. de Gennaro, A. Langella, Italian zeolitized rocks of technological interest, Miner. Deposita 31 (6) (1996) 452–472, http://dx.doi.org/10.1007/ BF00196127.

[46] M. de Gennaro, C. Colella, Use of thermal analysis for the evaluation of zeolite content in mixtures of hydrated phases, Thermochim. Acta 154 (2) (1989) 345–353, http://dx.doi.org/10.1016/0040-6031(89)85472-3. [47] M. Guadagno, The direct route gallery. Third contribution to the knowledge of the underground city, vol. 78, Atti R. Ist. Inc., Napoli, 1926, pp. 1–75 (in italian). [48] P. Nicotera, Contribution to the knowledge of the trachytic tuff of the hill of Vomero (Naples), Boll. Soc. dei Natur., Napoli, vol. 58, 1949, pp. 16–33 (in italian). [49] A. Pellegrino, Physical and mechanical properties of the volcanic soils of the Neapolitan area, Proc. VIII CNG, ESI, Cagliari, vol. 3, 1967, pp. 113–145 (in italian). [50] A. Evangelista, A. Pellegrino, Geotechnical characterization of some italian soft rocks, in: G. Barla (Ed.), Atti del 3 Ciclo di Conferenze di Meccanica ed Ingegneria delle Rocce (MIR), SGE, Padova, 2, 1990, pp 1–32. (in italian). [51] A. Colella, D. Calcaterra, P. Cappelletti, C. Di Benedetto, A. Langella, L. Papa, A. Perrotta, C. Scarpati, M. de Gennaro, The Neapolitan Yellow Tuff, in: M. de Gennaro, D. Calcaterra, A. Langella (Eds.), Le Pietre storiche della Campania: dall’oblio alla riscoperta, Luciano Editore, Napoli, 2013, pp. 129–154. (in italian). [52] B. Fitzner, Porosity properties and weathering behavior of natural stonesMethodology and examples, in: F. Zezza (Ed.), Proceedings of the C.U.M. 2nd Course ‘‘Stone materials in monuments: diagnosis and conservation”, Crete (Greece), University School of Monument Conservation, Bari, Crete (Greece), 1984, pp. 43–54. [53] D. Camuffo, Physical weathering of stones, Sci. Total Environ. 167 (1995) 1–14, http://dx.doi.org/10.1016/0048-9697(95)04565-I. [54] E.M. Winkler, Stone in Architecture: Properties, Durability, third ed., Springer, Berlin, 1997, http://dx.doi.org/10.1007/978-3-662-10070-7. [55] R.A. Kühnel, Driving forces of rock degradation, in: E. Galan, F. Zezza (Eds.), Protection and Conservation of the Cultural Heritage of the Mediterranean Cities, Balkema Publishers, Rotterdam, 2002, pp. 11–17. [56] Á. Török, Influence of fabric on the physical properties of limestones, in: S.K. Kourkoulis (Ed.), Fracture and Failure of Natural Building Stones, Springer, Dordrecht, 2006, pp. 487–495. [57] S. Siegesmund, H. Dürrast, Physical and mechanical properties of rocks, in: S. Siegesmund, R. Snethlage (Eds.), Stone in Architecture: Properties, Durability, fourth ed., Springer, Berlin, 2011, pp. 97–225, http://dx.doi.org/10.1007/9783-642-14475-2_3. [58] J. Ruedrich, S. Siegesmund, Salt and ice crystallization in porous sandstones, Environ. Geol. 52 (2) (2006) 225–249, http://dx.doi.org/10.1007/s00254-0060585-6. [59] R. Prˇikryl, T. Lokajícˇek, J. Svobodová, Z. Weishauptová, Experimental weathering of marlstone from Prˇední Kopanina (Czech Republic) - historical building stone of Prague, Build. Environ. 38 (2003) 1163–1171, http://dx.doi. org/10.1016/S0360-1323(03)00073-8. [60] D. Benavente, M.A. García del Cura, R. Fort, S. Ordóñez, Durability estimation of porous building stones from pore structure and strength, Eng. Geol. 74 (2004) 113–127, http://dx.doi.org/10.1016/j.enggeo.2004.03.005. [61] L.A. Rijniers, H.P. Huinink, L. Pel, K. Kopinga, Experimental evidence of crystallization pressure inside porous media, Phys. Rev. Lett. 94 (7) (2005), http://dx.doi.org/10.1103/PhysRevLett. 94.075503. epub 075503. [62] Á. Török, L.Z. Forgo, T. Vogt, S. Lobens, S. Siegesmund, T. Weiss, The influence of lithology and pore-size distribution on the durability of acid volcanic tuffs, Hungary, in: R. Prˇikryl, B.J. Smith (Eds.), Building stone decay: from diagnosis to conservation, Geological Society, vol. 271, Special Publication, London, 2007, pp. 251–260, http://dx.doi.org/10.1144/GSL.SP.2007.271.01.24. [63] E. Molina, G. Cultrone, E. Sebastián, F.J. Alonso, L. Carrizo, J. Esbert, O. Buj, The pore system of sedimentary rocks as a key factor in the durability of building materials, Eng. J. 118 (2011) 110–121, http://dx.doi.org/10.1016/j. enggeo.2011.01.008. [64] R. Prˇikryl, Some microstructural aspects of strength variation in rocks, Int. J. Rock Mech. Min. Sci. 38 (5) (2001) 671–682, http://dx.doi.org/10.1016/S13651609(01)00031-4. [65] B. Vásárhelyi, P. Ván, Influence of water content on the strength of rock, Eng. Geol. 84 (2006) 70–74, http://dx.doi.org/10.1016/j.enggeo.2005.11.011. [66] Á. Török, B. Vásárhelyi, The influence of fabric and water content on selected rock mechanical parameters of travertine, examples from Hungary, Eng. Geol. 115 (2010) 237–245, http://dx.doi.org/10.1016/j.enggeo.2010.01.005. [67] M.Y. Çelik, A. Ergül, The influence of the water saturation on the strength of volcanic tuffs used as building stones, Environ. Earth Sci. 74 (4) (2015) 3223– 3239, http://dx.doi.org/10.1007/s12665-015-4359-x. [68] A. Evangelista, S. Aversa, T.S. Pescatore, F. Pinto, Soft rocks in southern Italy and role of volcanic tuffs in the urbanization of Naples, in: Proceedings of the 2nd International Symposium on Hard Soils and Soft Rocks, Napoli, Italy, vol. 3, Balkema, the Netherlands, 2000, pp. 1243–1267. [69] A. Tugrul, The effect of weathering on pore geometry and compressive strength of selected rock types from Turkey, Eng. Geol. 75 (2004) 215–227, http://dx.doi.org/10.1016/j.enggeo.2004.05.008. [70] A. Langella, P. De Simone, D. Calcaterra, P. Cappelletti, M. de Gennaro, Evidence of the relationship occurring between zeolitization and lithification in the yellow facies of Campanian Ignimbrite (southern Italy), in: R. Aiello, E. Giordano, F. Testa (Eds.), Stud. Surf. Sci. Catal., vol. 142, 2002, pp. 1775–1782, http://dx.doi.org/10.1016/S0167-2991(02)80352-4. [71] A.B. Hawkins, Aspects of rock strength, Bull. Eng. Geol. Env. 57 (1) (1998) 17– 30, http://dx.doi.org/10.1007/s100640050017.

A. Colella et al. / Construction and Building Materials 136 (2017) 361–373 [72] D. Calcaterra, A. Langella, P. Cappelletti, V. Morra, R. de Gennaro, A. Colella, M. de Gennaro, Mineralogical and technological characterization of ‘‘Piperno”, a widely used building stone of Napoli (Italy), in: Proc. 6th International Congress on Applied Mineralogy, Balkema, Rotterdam, 2000, pp. 963–966. [73] D. Calcaterra, P. Cappelletti, A. Langella, V. Morra, R. de Gennaro, A. Colella, The building stones of the ancient centre of Naples (Italy): the Piperno from Phlegrean Fields. Contributions to the knowledge of a long-time used stone, J. Cult. Herit. 1 (4) (2000) 415–427, http://dx.doi.org/10.1016/S1296-2074(00) 01097-9. [74] D. Calcaterra, A. Langella, R. de Gennaro, M. de Gennaro, P. Cappelletti, Piperno from Campi Flegrei: a relevant stone in the historical and monumental heritage of Naples (Italy), Environ. Geol. 47 (3) (2005) 341–352, http://dx.doi. org/10.1007/s00254-004-1156-3. [75] D. Calcaterra, P. Cappelletti, M. de Gennaro, R. de Gennaro, F. de Sanctis, A. Flora, A. Langella, The rediscovery of an ancient exploitation site of Piperno, a valuable historical stone from Phlegraean Fields (Italy), in: R. Prˇikryl, B.J. Smith (Eds.), Building Stone Decay: From Diagnosis to Conservation, Geological Society, London, Special Publications, vol. 271(1), 2007, pp. 23– 31, http://dx.doi.org/10.1144/GSL.SP.2007.271.01.03. [76] G.F. Andriani, N. Walsh, Thermal properties and their influence on strength and deformability of carbonate rocks, in: G. Calvi, U. Zezza (Eds.), Proceedings of the International Congress ‘‘Quarry—Laboratory—Monument”, Pavia, 2000, pp. 81–90. [77] S. Siegesmund, K. Ullemeyer, T. Weiss, E.K. Tschegg, Physical weathering of marbles caused by anisotropic thermal expansion, Int. J. Earth Sci. 89 (1) (2000) 170–182, http://dx.doi.org/10.1007/s005310050324. [78] D. Benavente, G. Cultrone, M. Gómez-Heras, The combined influence of mineralogy, hydric and thermal properties in the durability of porous building stones, Eur. J. Min. 20 (2008) 673–685, http://dx.doi.org/10.1127/0935-1221/ 2008/0020-1850.

373

[79] Á. Török, M. Hajpál, Effect of temperature changes on the mineralogy and physical properties of sandstones. A laboratory study, Restor. Build. Monum. 11 (4) (2005) 1–8. [80] O. Marino, G. Mascolo, R. Cioffi, A. Colantuono, S. Dal Vecchio, Volcanic tuffs: chemical and physical deterioration mechanisms and restoration technique, L’Edilizia, vol. 9, 1991, pp. 523–535 (in italian). [81] J. Delgado, Rodrigues, Swelling behaviour of stones and its interest in conservation. An appraisal, Mater. Constr. 51 (263–264) (2001) 183–195, http://dx.doi.org/10.3989/mc.2001.v51.i263-264.363. [82] E. Steindlberger, Volcanic tuffs from Hesse (Germany) and their weathering behavior, Environ. Geol. 46 (3) (2004) 378–390, http://dx.doi.org/10.1007/ s00254-004-1039-7. [83] G.W. Scherer, I.J. Gonzalez, Characterization of swelling in clay-bearing stone, in: A.V. Turkington (Ed.), Stone decay in the architectural environment, Geological Society of America, Special paper 390, 2005, pp. 51–61, http://dx. doi.org/10.1130/0-8137-2390-6.51. [84] J. Ruedrich, T. Bartelsen, R. Dohrmann, S. Siegesmund, Moisture expansion as a deterioration factor for sandstone used in buildings, Environ. Earth Sci. 63 (7) (2011), http://dx.doi.org/10.1007/s12665-010-0767-0. 7–8, 1545–1564. [85] E. Sebastian, G. Cultrone, D. Benavente, L.L. Fernandez, K. Elert, C. Rodriguez, Navarro, Swelling damage in clay-rich sandstones used in the church of San Mateo in Tarifa (Spain), J. Cult. Herit. 9 (1) (2008) 66–76, http://dx.doi.org/ 10.1016/j.culher.2007.09.002. [86] R. López-Doncel, W. Wedekind, R. Dohrmann, S. Siegesmund, Moisture expansion associated to secondary porosity: an example of the Loseros Tuff of Guanajuato, Mexico. Environ. Earth Sci. 69 (4) (2013) 1189–1201, http://dx. doi.org/10.1007/s12665-012-1781-1. [87] W. Wedekind, R. López-Doncel, R. Dohrmann, M. Kocher, S. Siegesmund, Weathering of volcanic tuff rocks caused by moisture expansion, Environ. Earth Sci. 69 (4) (2013) 1203–1224, http://dx.doi.org/10.1007/s12665-0122158-1.