Monument Future

Methods

The archaeological site of Mitla, the church of Santo Domingo de Guzmán and the Oaxaca Cathedral were mapped for lithology, intensity of damage and weathering features (Fig. 2). For most walls a representative area was selected, and each rock was mapped individually to make a semi-quantitative analysis.

In accordance to the German industrial norms, several laboratory tests were conducted for the analysis of the petrophysical and moisture parameters as porosity, density, capillary water absorption, water vapor diffusion, hygroscopic water sorption, 55ultrasonic velocity, rebound hardness and tensile strength. For the analysis of weathering properties, tests on the thermal expansion, hydric expansion and salt bursting were performed. The petrophysical properties were analyzed parallel (X-direction) and perpendicular to the bedding plane (Z-direction). A detailed description of the laboratory analysis can be found in (Siegesmund and Dürrast 2011).

Figure 2: Mapping in Oaxaca and Mitla. a): Lithological map of Santo Domingo de Guzmán, b): Weathering features at the eastern side of the Oaxaca Cathedral, c): Intensity of damage at the southern side of ‘el palacio’ at the archaeological site of Mitla.

Results

Table 1 summarizes the test results for the petrophysical and moisture properties as well as the weathering properties of all samples.

Mapping

The historical buildings in Oaxaca were mostly built with CVO. Delicate parts like ornaments and window sills were built with the softer CAO. CVE was used as a replacement for CVO during different phases of restoration (Fig. 2a).

The most important weathering features in Oaxaca are loss of components, loss of matrix and bursting. The base parts of the buildings often show scaling and flaking accompanied by salt crust formation. Upper wall parts and protruding areas are strongly affected by moisture and form a characteristic organic black crust (Fig. 2b). CVO is dominated by loss of components when pumice clasts dissolved. CAO and CVE are strongly affected by fracturing, when in contact with water (Kück et al. 2020b). CRO and CRE are the only samples which show bedding parallel scaling.

In Mitla most of the building stones are original and mostly in good condition. The southern and eastern side of the buildings have the highest damage intensities linked to high temperature changes during the day of up to 20 °C (Fig. 2c). The northern side experiences a lot of moisture, but the temperature difference during the day is low (3 °C). The western side is the least weathered side.

Petrophysical and moisture properties

The effective porosity of all analyzed tuffs varies from 12 % (MG) to 41 % (CAO). The Mitla tuffs show the lowest values (12–15 %). The sample with the highest matrix density is MR with 2.6 g/cm3, while MG has the lowest density with 2.2 g/cm3 (Tab. 1). The saturation degree S of the samples ranges from 0.59 (MG) to 0.92 (MR, CVO). After Hirschwald (1912) most samples are weathering and frost resistant except MR and CVO. The pore size distributions were measured for all samples (Fig. 3). It is noticeable, that CVO and MR have almost no capillary pores, but a high share of micropores.

Figure 3: Comparison of the ratio of micropores and capillary pores.

The samples from Mitla show the lowest capillary water absorption (0.7 kg/m2√h–1.7kg/m2√h). The samples from Oaxaca show moderate w-values ranging from 1.0–2.8 kg/m2√h (CV ) to 6.4–8.5 kg/m2√h (CAO). CVO has the highest water vapor diffusion resistance factor μ with 21.9, followed by MR with 19.0. The sample with the lowest water vapor diffusion resistance is CAO with 7.4. Hygroscopic water sorption was measured from 20–95 % rh. The highest sorption value was measured for CVE with 7.75 wt.-% at 95 % rh. The lowest sorption at 95 % rh was measured for CRO with 1.74 wt.-%. Between 20 and 75 % rh all samples show a linear weight increase with varying gradient. Above 75 % rh the curves show a sudden increase in slope. The highest ultrasonic velocities were measured for MR (3.7–3.8 km/s) and CVO (3.2–3.6 km/s). The lowest ultrasonic velocity was measured for CRE (1.6–2.3 km/s).

The tensile strength of all samples varies between 1.3–1.7 MPa (CRE) and 6.4–6.9 MPa (CVO). MR has the highest surface hardness with 528–604 HL followed by CVO (603–632 HL). The lowest surface hardness was measured for CAO with 330–337 HL.56

Table 1: Petrophysical properties and weathering behaviour of the studied samples.

Weathering properties

Thermal dilatation was tested in two heating cycles each in both a dry and wet environment. The samples CVO, CVE and CAE show pronounced differences between the first and second heating cycle (Fig. 4). All samples expand during heating and then shrink again during cooling. The samples from Oaxaca and Etla have negative residual strain, especially after the second heating cycle. At about 70–80 °C the samples shrink significantly. The samples from Mitla reach their initial state after cooling and the first and second heating cycle are similar. CAE has the highest thermal expansion with 0.78–0.87 mm/m. The tuff varieties from Oaxaca and Etla have the lowest thermal expansion (0.17–0.22 mm/m).

After two dry heating cycles the samples were saturated with water and heated in two wet heating cycles. After the first hour in a wet environment the samples already showed intense hydric expansion with a maximum of more than 2 mm/m. After that first intense expansion the samples 57underwent a thermohydric dilatation. For thermohydric dilatation CRO and CRE show much higher anisotropies than for just thermal dilatation. The highest thermohydric expansion was measured for CVO (2.90 mm/m) and CVE (1.98–2.25 mm/m). The samples from Mitla have the lowest thermohydric expansion.

Figure 4: Thermal and thermohydric expansion.

The second dry cycle shows a significant decrease of residual strain, with negative values representing thermal contraction. After the first wet heating cycle the samples show a decrease in residual strain. Especially CVO, CVE and CAE show this feature. CAO and CAE do not recover from contraction after the second drying cycle. The Mitla samples are rather unaffected by both thermal and thermohydric dilatation.

Figure 5: Maximum hydric expansion values.

Figure 5 shows the hydric expansion of all tested samples. CVO expands the most of all samples, when in contact with water. The maximal hydric expansion is reached by CVO in the Z-direction with 2.36 mm/m. The smallest hydric expansion was measured for MR with 0.04 mm/m in the X-direction. The sample with the highest anisotropy is MR with 73 %. Most samples show higher hydric expansion in the Z-direction, except for CAO, CAE and MG. All samples reach maximal hydric expansion already during the first hours.

The effect of salt weathering on all samples is presented in Figure 6. Most of the samples show a weight increase at the beginning of the salt bursting test. The most resistant sample is MR, while CRE, CRO and CVO show low resistance towards salt weathering. Table 1 shows the number of salt cycles until at least 30 % weight loss is reached.

Discussion and conclusion

The combination of the mapping results and the experimental tests allow an evaluation of the rocks. The samples from Mitla are in general very resistant to most forms of weathering. They have low porosities and low w-values with low hydric 58and thermohydric dilatation. Both MG and MR are relatively resistant to salt weathering compared to the other analyzed tuff samples.

Figure 6: Photographic documentation of the salt bursting test.

The pore space has a very important impact on the process of salt weathering. A general dependence between the share of micropores and the number of salt cycles could be observed in the samples. MR and CVO are the strongest samples with high surface hardness, ultrasonic velocities, tensile strength and low porosity with almost no capillary pores present. CVO is not very resistant to salt weathering and shows the highest hydric expansion and strong thermohydric dilatation with non-reversible changes due to the dehydration of clay minerals. This makes CVO less suitable as a building stone. Nevertheless, CVE is even more affected by moisture than CVO. With the highest sorption, high hydric expansion and low tensile strength as well as low ultrasonic velocity CVE is strongly affected by loss of components and fracturing/bursting. CAO and CAE are less affected by component loss and show strong back weathering and fracturing due to a high amount of clay minerals, which makes the yellow tuff varieties rather soft samples. The red tuffs (CRO and CRE) are the only ones which show anisotropic behavior in the weathering features that correlate with relatively high anisotropic properties.

The results suggest, that moisture and thermal changes are the most severe damaging factors for tuffs from Oaxaca. The southern and eastern sides of most buildings are the ones with the highest temperature changes, and therefore are weathered the strongest. The northern side usually experiences a lot of moisture, while the western side is in general the least weathered building side. The tuff varieties from Mitla are very suitable building stones and have remained in the buildings for a long time. The green tuff from Oaxaca (CVO) is a historical building stone of good quality but needs to be protected from moisture. The replacement stone from Etla (CVE), which is used today, is a suitable building rock but is certainly of a lower quality than CVO. Therefore, conservational treatment should be considered for CVE.

Acknowledgements

This work was supported by the German Research Foundation (Si-438/52-1) and the German Federal Environmental Foundation (AZ20017/481). We would like to thank J. L. Noria Sánchez, A. E. Andrade Cuautle and M. Á. González for the support of the on site work. For the laboratory support and helpful comments we thank K. Wemmer, J.Menningen, W. Wedekind and C. Gross.

References

Bernal, I. (1963), Otra tumba cruciforme de Mitla, Estudios de cultura náhuatl, Vol. 4, pp. 223–232.

Blanton, R., Feinman, G., Kowalewski, S. and Nicholas, L. (1999), Ancient Oaxaca, Vol. 2, Cambridge University Press.

Casanova, J. P., and Pino, L., (2004) Templos históricos de Oaxaca, 1. ed.: Oaxaca de Juárez, H. Ayuntamiento de Oaxaca de Juárez, H. Ayuntamiento Oaxaca de Juárez (Series), Volume 2.

García, N. M. R. (2016), Mitla: Su desarrollo cultural e importancia regional, Fondo de Cultura Económica.

Hirschwald, J. (1912), Die Prüfung der natürlichen Bausteine auf ihre Wetterbeständigkeit, W. Ernst & Sohn, Berlin.

Kück, A., Pötzl, C., López-Doncel, R. A., Dohrmann, R. and Siegesmund, S. (2020b), Effects of zeolites and swellable clay minerals on water related properties and thermal dilatation in volcanic tuff rocks, in Siegesmund, S. and Middendorf, B. (Eds.), Monument future: Decay and conservation of stone, Göttingen, Kassel, Mitteldeutscher Verlag, Halle.

Siegesmund, S. and Dürrast, H. (2011), Physical and mechanical properties of rocks, in Siegesmund, S. and Snethlage, R. (Eds.), Stone in architecture, Springer, pp. 97–225.

Urquiaga, J. (2000), La restauración del ex Convento de Santo Domingo, Oaxaca, Consejo Nacional para la Cultura y la Artes, México.

SUSTAINABLE LIME RESTORATION MORTARS: PHYSICAL PROPERTIES AND DURABILITY ASSESSMENT

José Diaz, Beatriz Menéndez

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

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

VOLUME I AND VOLUME II. MITTELDEUTSCHER VERLAG 2020.

Geosciences and Environment Cergy, Cergy-Pontoise University, France, jose.diaz-basteris@u-cergy.fr, beatriz.menendez@u-cergy.fr

Abstract

In order to improve the ecological footprint of the restauration mortars while keeping their efficiency, we have tested several combinations of natural hydraulic lime and air lime with natural aggregates and additives. Repair mortars should be compatible with the support (stones, bricks) and protect original materials from environmental agents; aesthetical and historic aspects must not be neglected. We started by “simple” mixtures of lime and natural siliceous sand or crushed calcareous limestone. Aggregates granulometry was 0/2 mm in all the cases. More “complex” mortars were also prepared by incorporating the additives pinecone resin, semimilled cones of pine, milled glass waste, chamotte (brick production residue).

Different physical properties have been measured in all mortars: porosity, density, capillarity absorption, mechanical strength (flexural and compression), P and S waves velocities from which dynamic Young’s modulus and Poisson’s ratio were inferred. Durability of mortars has been estimated by salt crystallization and frost/thaw cycles.

Keywords: Restoration mortars, sustainable Cultural Heritage conservation, Building materials, Durability

Introduction

Restoration mortar designates a group of products made to repair a damaged masonry. They are used to rebuild the lost parts of buildings, monuments or sculptures that have disappeared and to make replicas of architectonic or sculptural elements of high cultural value, placed inside buildings or museums to ensure their conservation. Mortars for restorations must be carefully selected before application. They have to be able to interact with the ancient stone and match in a chemical, physical and mechanical way with the pre-existing materials (Rota Rossi Doria, 1986; Torney, 2016). The choice of the mortar must take into account the support properties and the environment in which the mortar will be applied. Nowadays, in the development of new commercial restoration mortars, their environmental impact and their durability in future conditions are not always considered.

Our work aims at proposing solutions to produce more ecological and durable mortars. This research will allow the development of a series of new restoration mortars which design will be done in a “circular economy” philosophy.

To develop new sustainable formulas, we tested different additives available in Paris region, with a low CO2 footprint and capable to improve the 60mechanical properties of resulting mortars. Inorganic additives such as crushed bricks, crushed glass and slags have been identified in historical mortars. The use of recycled materials is considered as a zero CO2 emission contribution to the mortar footprint; only the energy used to transform these materials: grinding or cleaning is considered in the total environmental balance. For this reason we included recycled additives in our formulations. Crashed bricks (Aalil et al 2019) and waste glass powder (GWP) have been used in construction works because they improves the durability and the compressive strength of lime mortars (Carsana et al. 2014). GWP provides a pozzolanic behaviour to lime mortar (Edwards et al. 2007).

Many studies have been done about the influence of kind of binder, composition and particle size of aggregates on the physical and chemical properties of mortars. For example, the use of limestone sand as aggregate is recommended to ensure a better compatibility with natural stones (Szemerey-Kiss et al. 2011).

In ancient cultures like Mayans and Romans, organic additives were used to improve the properties of the mortars (Villaseñor et al. 2009; Ordóñez et al. 2019). For example, the tannin and sugar content of the pinecones ensured water resistance and plasticity of mortars (Rampazzi et al 2016, Arcolao 1998). In restoration, it is very important to dispose of mortars with different ranges of physical and mechanical properties using natural additives and without using any cement. (Russlan et al. 2018).

Materials and Methods

The amount of raw materials is limited and must be preserved and managed with caution. One important aspect of the presented work is to save natural raw materials when formulating restauration mortars. In order to increase the sustainability of the restoration mortar industry, the use of waste materials will be enhanced.

Different mixtures were prepared in order to determine the influence of components in the physical properties of harder mortars. Two binders, four types of sand (aggregates) and four additives were employed in the fabrication of studied mortars. The binders are: aerial lime (Cl90) and hydraulic lime (NHL5) both provided by Socli company. The aggregates are: silica sand (Leroy Merlin), calcareous sand provided by Rocamat company, silico-calcareous sand (Italcementi group) and fine silica sand (Sibelco). Four different additives were used: chamotte provided by Briqueterie d’Allone, ground glass provided by the Fédération du Verre, ground pine cone and water/oil solution of pine cone.

Mortars were prepared in the laboratory at average temperature of 24 °C and relative humidity of 40 %. The preparation process consists in several steps: 1) sands were dried at 60 °C during 24 h, 2) weighing of the components, 3) dry mixing using an electric mixer (Rubimix 9), 4) addition of water and mechanical mixing for 3 minutes, 5) mortars are moulded in prismatic casts, 40 mm × 40 mm × 160 mm and placed in hermetic plastic boxes for 7 days to preserve 90 % humidity, 6) samples are unmoulded and put in a humidity chamber at 65 % for 21 days following the standard EN 1015-2. Finally, the samples were stored at laboratory conditions.

The different formulations obtained are presented in Table 1. To name the mortars we use the initials of the components: H for hydraulic lime, A for aerial lime, C for calcareous sand, S for silica sand, F for silico-calcareous sand, D for fine silica sand, 61G for waste glass powder, B for chamotte (crushed brick), P for pinecone and R for the resin of the pinecone.

Figure 1: Mortars samples at 180 days. The different formulations obtained are presented in Table 1.

The first mortar prepared was the HFD after evaluation it was decided to increase the amount of the binder, to increase the resistance of 20 %. The mortars HS and AS are performed to compare binders, the mortar HB is prepared to test the additive, the mortar HSD is prepared to test the effect of granulometry to increase de resistance.

The HSC and HSCR mortars start from the same base with 30 % of binder but one is prepared with water and the other with the pine resin solution, mortar HC was prepared to test the calcareous sand, and HSG for test the WPG.In future work more formulas will be prepared to make comparisons.

Several properties have been measured in the harden mortars (90 or 180 days old). Porosity and density have been measured by the triple weight method following the standard EN 1936. Compressive and flexural strength were obtained in 4 × 4 × 16 cm³ samples, according to EN 1015-11 standard. Capillarity water absorption tests were performed following the standard EN 1015-18. Durability of the samples was estimated by salt crystallisation and frost resistance tests following standards EN 12371 and EN 12370. Dynamic Young’s modulus (E) and Poisson’s coefficient (n) from P and S wave velocities using the next equations (Baron 2007):

In order to validate the obtained formulations, their physical properties have been compared with three commercial products (Altarpierre, Artopierre and Lithomex TM) previously studed in our laboratory and largely employed in France and others parts of the world.