CC BY-NC-ND
0Available online at www.sciencedirect.com
SciVerse ScienceDirect Procedia
Engineering
Procedia Engineering 57 (2013) 565 - 571 =
www.elsevier.com/locate/procedia
11th International Conference on Modern Building Materials, Structures and Techniques,
MBMST 2013
The Effect of Composition of Roman Cement Repair Mortars on Their Salt Crystallization Resistance and Adhesion
Anna Klisinska-Kopacza *, Renata Tislovab
aNational Museum in Krakow, al. 3 Maja 1, PL-30 062, Krakow, Poland b Faculty of Restoration, University of Pardubice, Jiraskova 3, CZ-57001, Litomysl, The Czech Republic
Abstract
Crystallization of soluble salts in porous networks is a major source of decay for mortars used in historical buildings. The optimum formulation and application of Roman cement mortars which could produce compatible and durable repairs of the original substrates was studied. Measurements were performed with the aim of determining the pore size distribution, the hydric parameters as well as the salt crystallization resistance of the mortars. The adhesive strength of the repair materials laid on historic substrates was also determined. The results of the crystallization tests show that repair Roman cement mortars with hydric parameters close to those of the historic substrates, though different pore size distributions, have related salt crystallization resistance.
© 2013 The Authors. Published by Elsevier Ltd.
Selection and peer-review under responsibility of the Vilnius Gediminas Technical University Keywords: Roman cement, historic mortars, pore structure, salt crystallisation, adhesion, mercury porosimetry.
1. Introduction
Highly hydraulic bindurr, known ar natural nr Roman cumuntr, wuru kuy materials fnr thu ucnnnmic and uary manufacture nf rundurr and cart archituctural dutailr fnr thu uxturinrr nf buildingr during thu ninutuunth and uarly twuntiuth cuntury [1]. Thu grnwing availability nf Rnman cumuntr fnr thu cnnrurvatinn fiuld har rairud impnrtant quurtinnr abnut thu cnmpatibility butwuun hirtnric and rupair mnrtarr. Cnmpatibility ir brnadly dufinud ar thu capacity nf thu rupair mnrtar tn inturact with thu nriginal hirtnric maturial withnut cauring any damagu, diructly nr indiructly [2]. Applicatinn nf an nld rucipu nr mimicking nf prnpurtiur nf nriginal mnrtar givur nn guarantuu nf maturial cnmpatibility bucauru bnth maturialr uvnlvu in cnurru nf timu. Ar thu prnpurtiur nf thu rupair mnrtarr can vary duu tn thu inturactinn nf frurh mnrtarr with thu rubrtratu during harduning, ruvural cnmbinatinnr nf hirtnric rubrtratur and rupair mnrtarr rhnuld bu rtudiud.
Cryrtallizatinn nf rnlublu raltr in pnrnur nutwnrkr ir a majnr rnurcu nf ducay fnr mnrtarr urud in hirtnrical buildingr. Thu principlu nf thir ducay ir rnughly uxplainud ar raltr cryrtallizing in thu pnrnur nutwnrkr and cauring rtrurr inridu thu pnrur [3], [4]. Thir rtrurr ir rurpnnriblu fnr thu inturnal cracking nf thu mnrtar. Thu typu nf damagu dupundr nn thu typu nf ralt ar wull ar nn thu characturirticr nf thu maturialr. Thu mnrtarr and rtnnur that aru mnrt runritivu tn ralt ducay aru gunurally thnru with high pnrnrity and high ratu nf watur uvapnratinn. Thu ducay nf natural rtnnur caurud by ralt cryrtallizatinn har buun widuly dircurrud [5-7]. Hnwuvur, in thu cnntuxt nf pnrnur building maturialr, thu cryrtallizatinn nf rnlublu raltr in Rnman cumunt mnrtarr har nuvur buun durcribud.
Onu nf thu principal aimr nf thir rtudy war tn iduntify rciuntifically-rnund crituria fnr thu nptimum fnrmulatinn and applicatinn nf Rnman cumunt mnrtarr that cnuld prnducu cnmpatiblu and durablu rupairr nf thu nriginal rubrtratur. Gnnd adhurinn butwuun thu nriginal and thu rupair maturial ir annthur vital cnnditinn nucurrary fnr thu prnpur uxucutinn nf cnnrurvatinn wnrk. Othurwiru, any factnr ruducing thu adhurivu rtrungth nf thu rupair can maku a gnnd rupair maturial
* Corresponding author.
E-mail address: aaklisinska@muzeum.krakow.pl; brenata.tislova@upce.cz
ELSEVIER
1877-7058 © 2013 The Authors. Published by Elsevier Ltd.
Selection and peer-review under responsibility of the Vilnius Gediminas Technical University doi: 10.1016/j.proeng.2013.04.072
incompatible. Only a few publications report on the adhesion strength between repair material and concrete substrate [8], [9]. The pore structure of the mortars is one of the main properties responsible for their compatibility as it greatly influences water adsorption, shrinkage cracking [10] and strength. The mismatch in the pore structure at the substrate-mortar interface may lead to a zone of damage due to salt crystallization affecting the adhesive strength of a repair, and thus compromising its final compatibility. Repair mortars should be less durable than the historic substrate; their role should ultimately be sacrificial. Different wetting and drying rates may also adversely affect the appearance of the repair, whereas a repair mortar with too high strength imposes stress on weaker masonry or stuccoes.
This paper describes the influence of mortar composition on their salt crystallization resistance. Mortars with different composition may enhance water and salt solution transport resulting in different types of damage. To establish the compatibility of historic and repair mortars, salt crystallization resistance tests were performed on sandwich samples composed of different substrates covered with historic and repair Roman cement mortars. Sodium sulphate was the agent selected for this study since it is widely known as the most destructing salt affecting porous building materials and its damaging mechanisms are well known [11-13]. Measurements of the adhesive strength of repair materials laid on historic substrates were also performed while good adhesion of repair mortars was another criterion for the compatibility assessment. In the case of poor adhesion, local delamination between the mortar and the substrate layer may occur resulting in crystallization of salt in areas without strict adherence.
2. Materials and methods
The formulation of the restoration mortars used in this study replicated some representative groups of historic mortars. Two sets of repair mortars were prepared from Folwark Roman cement (Poland) by mixing the cement with the aggregate (a/c = 0.5 by weight). Quartz sand (d = 0.25-0.5 mm) was used as the aggregate. The mortars were produced at water-to-cement ratios varying from 0.6 to 1.0. Due to the quick setting of the cement, citric acid was added as a retarder at 0.3 wt.% (relative to dry cement). Mortars were casted into silicone molds. Masonry samples with two different types of joints were considered. The first set consisted of newly prepared mortars specimens of 5x5x2 cm. The second set of specimens consisted of a fired-clay brick substrate or historic substrates (Imperial Palace Hofburg, Vienna, Austria; core of a casting) with laid Roman cement repair mortars. During preparation of the sandwich samples, the substrates were pre-wetted by soaking in water. Specimens were about 5x5x5 cm3 with the plaster layer being approximately 2 cm thick. The materials were then de-molded after setting and stored in humid air at 100% relative humidity (RH) conditions for 3 months to ensure wet curing conditions. The influence of different curing conditions on hydration has been previously described [14]. The lack of moisture in the external conditions may result in incomplete hydration and affect the microstructure and strength of the mortars. In real world conditions, ideal wet-air curing of mortars is rarely possible in the course of practical work on the façade. The characteristics of the original marl feedstock, the calcinations conditions, the oxide and mineralogical compositions of the cements, as well as their strength and porosity development in the progress of hydration have already been reported [15].
The pore structure of the mortar samples was determined using a Poremaster mercury intrusion porosimeter (quantachrome), which allows studying pore sizes in the range of 440-0.0035 ^m. After the predetermined curing period, the specimens were immediately soaked in acetone for 24 h to stop the hydration of the cementitious materials [16], [17]. Afterwards, they were placed in a rotary vacuum evaporator flask at 20 °C for 4 h to remove the acetone and were finally allowed to dry.
Salt crystallization cycling allows determining the resistance of mortars to salt attack. Salt crystallization cycles were performed according to the procedure given by ASTM C88-05 [18]. Mortar/brick and mortar/historic substrate specimens of 100 ± 20 g were used for testing. Prior to each cycle, specimens were weighed. The samples were immersed in a 14% sodium sulphate decahydrate (Na2S04-10H20) solution for 4 h. The density of this solution at 20 °C is 1055 kg/m3. They were then removed from the solution and left to dry in an drier at 60 °C for 8 h and later left to cool to room temperature for 6 h. Their individual masses were noted before continuing the soaking in the salt solution. Sample cubes were subjected to 15 cycles except in the event of failure or disintegration of the cubes occurring before completion.
Hydric tests including the determination of the free water absorption (AF), the saturation coefficient (S) and the water absorption by capillarity (C) were performed to characterise the parameters associated with fluid uptake and transport inside the pores according to the procedure recommended by PN-EN 1015-18 [19]. The calculations were performed using the procedure described previously [20].
The adhesion of repair mortars has been studied on composite samples, which were prepared by applying a repair mortar onto historic Roman cement substrates and bricks. Pull-off strength measurements are recommended for the assessment of the adhesive strength between concrete substrates and repair materials. The pull-off test was carried out with a digital
apparatus PosiTest pull-off adhesion tester. Steel disks of 50 mm in diameter were bonded to the top of each repair mortar with epoxy glue. Pull-off tests were performed according to the procedure recommended by ASTM D7234-12 [21].
3. Results and discussion
Porosity and pore size distribution are important parameters used for describing a hydrated cement system when the physico-chemical behaviour and the durability of repair mortars is considered. Figs. 1 and 2 compare the differential volume of intruded mercury as a function of pore diameter for the Roman cement mortars prepared using water-to-cement ratios (w/c) of 0.6 and 1.0, studied after aging for 1 day and 3 months, respectively.
-w/c - 0 ,6
----w/c = 1.0 i
1 \ ' 'V M/'
yy \ \
y \ »,
1E-3 0,01 0,1 1 10 100
Pore diameter ( rim )
Fig. 1. Differential volume of intruded mercury vs pore diameter for mortars with w/c ratio of 0.6 and 1.0 cured for 1 day
These repair mortars were prepared using typical historic mortar formulations, which were previously studied by the authors [22]. The mortars with less amount of water represent the typical formulation of the cast mortars while the latter with 1.0 w/c is rather close to the rendering mortar. A curing time of 1 day represents fresh or low hydrated mortars; 3 months of hydration at ideal moist curing conditions is considered sufficient for complete hydration.
Fig. 2. Differential volume of intruded mercury vs pore diameter for mortars with w/c ratio of 0.6 and 1.0 cured for 3 months
After studying the porous structure of Roman cement repair mortars, the unimodal peak at 0.7 (m was observed for the 1-day old specimen prepared at the w/c ratio of 0.6. With the increase in water content, the bimodal pore structure can be observed. The initial peak shifts to higher pore diameters with a maximum at 1.9 (m. The second porosity region, similar as the one found in 0.6 w/c mortars, can be noticed with a pore diameter of 0.3 (m. For samples cured for 3 months, the threshold pore diameters are reduced to 0.01-0.03 (m. This shift is the result of the progress of hydration and the closing of the broad pore diameters by the growth of the C-S-H gel. However, the mercury intrusion curves point to the presence of
larger pores of 1.20 ^m in diameter for samples prepared at w/c ratios of 1.0. This observation can be interpreted in terms of an insufficient amount of C-S-H gel required to fill the large pores present in the structure due to an excess amount of water.
Similar pore-size regions were observed in historic Roman cement mortars with pore diameter values that were dependent on the initial water to cement ratio (w/c) or curing conditions. The finest pores, with diameters below 0.2 ^m, were present within the hardened aged Roman cement matrix. Pore diameters shift to smaller values in the case of wellhydrated and matured Roman cement mortars, which were exposed to wet-curing conditions over a long period of time (e.g. stucco, cast ornaments).
Larger 'air' pores, with diameters between 0.2-2 ^m, are due to the evaporation of excess unbound water and restricted hydration. They were characteristic of materials in which the hydration process was interrupted by the evaporation of water. Pores larger than 2 ^m were induced by shrinkage drying and mortar weathering [22].
The porous structure of fired-clay brick has a very close pore size distribution relative to that of historic Roman cement mortar. The only difference lies in the region of macropores whose maximum is shifted to 3.0 ^m showing a sharp peak with very narrow distribution (Fig. 3).
Fig. 3. Differential volume of intruded mercury vs pore diameter for a Roman cement historic mortar and a fired-clay brick
The weathering processes greatly depends on the porous structure and the water circulation inside the materials. A comparison of pore and hydric parameters of brick, historic and repair Roman cement mortars after 3 months of hydration is shown in Table 1. These results are complemented by the water absorption curves, which show the water behavior due to capillary action of the repair mortars and both substrates (Fig. 4).
Table 1. Porosity and hydric parameters of historic substrates and Roman cement repair mortars hydrated for 3 months.
Sample Total open porosity [%] Pore diameter [pm] Free water absorption, Af [%wt] Saturation coefficient, S [% v/v]
Roman cement repair mortar w/c = 0.6 30 0.01-0.03 22 42
Roman cement repair mortar w/c = 1.0 52 0.01-0.03 1.20 27 45
Roman cement historic substrate 22 0.03-0.1 1.0 0.02-0.04 3.0 20 36
Fired-clay brick substrate 23 17 29
The hydric parameters obtained for the Roman cement repair mortar with w/c ratio of 0.6 are only slightly higher from the ones obtained for the Roman cement historic substrate. The repair mortars prepared with higher w/c ratio possess more open spaces inside the matrix. This results in higher free water absorption compared to the repair mortars with lower w/c ratio as well as to that of obtained for the two original substrates. The water absorption by capillarity shows considerable water uptake of both repair mortars. This is mainly due to their high percentage of very small pores and their total porosity values.
Pore diameter {|i-n )
Fig. 4. Water absorption by capillarity for a fired-clay brick, a historic mortar substrate and a Roman cement repair mortar with w/c ratio of 0.6 and 1.0, cured for 3 months
The fired-clay brick and historic Roman cement substrate has lower water absorption by capillarity due to the lower total porosity that can be filled by water. Another general observation can be made from water absorption measurements - both repair Roman cement mortars and both substrates show fast transport of the water to the surface which determines the condition for crystallization of the salts on the surface of the materials. The small differences in the water absorption rate between the materials can be similarly related to the amount of capillary pores, which are active in water transport. It can be observed that the smaller the pores, the greater the capillary absorption.
The results of salt crystallization resistance of Roman cement repair mortars with varying w/c ratios for a fired-clay brick and a historic mortar substrate during 15 cycles of artificial aging in saturated sulphate solution are presented in Fig. 5. Analysis concerning the resistance to sulphates revealed that Roman cement repair mortars with w/c ratio of 1.0 collapse prematurely, with total destruction at the 8 cycle. In contrast, for the repair mortars with w/c ratio of 0.6, the deterioration occurred at a much slower rate. For repair mortars with lower w/c ratio, the first signs of damage were visible at 8 cycles and total destruction was noted at the 14 cycle similar to that obtained for the Roman cement historic substrate.
40 -, 3020100-10.20-30-40-
a repair mortar w/c - 0,6
o repair mortar w/c =1,0
■ fired-ctay brick
* historic mortar
O-O A ,
t > « É
4 e $ 10 12
Cycles completed
Fig. 5. Loss of mass (%) in relation to the initial mass of Roman cement repair mortars with w/c ratio of 0.6 and 1.0 cured for 3 months, a fired-clay brick and a historic mortar substrate during 15 cycles of artificial aging in saturated sulphate solution
For a porous materials such as stone, particular attention has been placed on the relationship between pore radius distribution and the resistance of the material to mechanical damage caused by salt crystallization. According to Angeli et al. [23], crystal growth of salts is favored in large pores because the chemical potential of crystallites depends inversely on their radii. This is due to the contribution of crystal surface energy, which increases when the crystal size decreases. Crystal growth in the smaller pores can start only after the larger pores are completely filled. At this moment, the increasing pressure of the crystals on the pore walls renders the chemical potential of the crystals equal in both classes of pores. According to Lanas et al. [24] and Cavdar and Yetgin [25], salt crystallization and mechanical stress are the most effective causes of degradation of lime-based mortars used in renders. Decay due to salt attack is most noticeable in mortars with
high porosity and low strength. In the case of Roman cement repair mortars, a similar situation can be noted. For Roman cement repair mortars with bimodal pore size distribution, the salt crystallization resistance is less compared to mortars with unimodal pore size distribution. Also, the higher amount of absorbed water and salt solution, as shown in previous experiments, can contribute to a more pronounced degradation pathway of the repair mortars.
Most crystallization tests reported in the literature are meant to assess the resistance/durability of certain materials to salt crystallization. Accordingly, they are performed on specimens made of a single material (concrete, mortar, brick, stone, etc.). Plasters and renders do not work independently of the substrate. Therefore, in the case of incompatibility of these two materials, their detachment is already evidence of damage, even if no other deterioration occurs in the plaster/render or in the substrate materials. The strength of adhesion profiles of repair mortars with w/c ratio of 0.6 and 1.0 cured for 3 months laid on a water-soaked brick and a Roman cement historic substrate is shown in Fig. 6.
Hydration time (days)
Fig. 6. Adhesion of repair mortars with w/c ratio 0.6 and 1.0 cured for 3 months laid on a water-soaked brick and a Roman cement historic substrate
A clear correlation between adhesion and curing time, equivalent to the degree of hydration, is observed. The adhesion increases significantly during the first 14 days of hydration when a pull-off strength level of 1.3 MPa is attained for the moist-cured specimen. The parameter further increases during the following 3 months, attaining approximately 2.0 MPa for the sample with a w/c ratio of 0.6 laid on the historic Roman cement substrate. The influence of mortars composition on the pull-off strength values can be clearly seen. Increasing the w/c ratio in the repair mortars lowers their adhesion. Mortars with w/c ratio of 1.0 exhibit comparatively lower values. However, the profiles of the pull-off strength development are the same for both mortars composition. For samples laid on fired-clay brick, independent of repair mortars composition, the strength of adhesion is very similar and also close to the values obtained for repair mortars with a w/c ratio of 1.0 laid on a historic substrate. The results show that mortars with higher porosity and water uptake are more favoured and cause harmful decay. In particular, layer detachment can cause a lack of adhesion between the mortar surface and the brick or the historic substrate.
4. Conclusions
In general, Roman cement stuccoes are in an excellent state of preservation in spite of their usual exposure to polluted urban environments for more than a century. When restoring these materials, particular attention must be paid to an adequate choice of repair mortars, which are critical to the success of the treatment. Compatibility between new repair mortars and the original components of the masonry is highly desirable. It is widely accepted that the problems introduced by replacement mortars are highly complex. However, when considering their resistance and durability to water and water soluble salt action, porosity structure is a crucial parameter that influences the performance of the repair mortar from the point of view of their composition and properties.
In this study, two types of Roman cement repair materials have been studied relative to their original substrates. The mortars differ in water content, which was found to have an influence on their porosity. In the repair mortars, unimodal distribution of pore sizes was observed for low water to cement ratios. With an increase in the water to cement ratios a development of a bimodal pore structure was noticed.
Even though, they were found to develop pore structures close to those of historic substrates which is very well reflected in transport of water and water salt solutions and their resistance to water soluble salts action. In the case of Roman cement repair mortars, a decay due to salt attack is most noticeable in mortars with high porosity. For Roman cement repair mortars
with bimodal pore size distribution, the salt crystallization resistance is less compared to mortars with unimodal pore size distribution. Therefore, the above considerations led to the conclusion that they are in broad terms compatible with historic masonry or stuccoes.
However, the porosity of the repair materials can be controlled by a careful manipulation of the water-to-cement ratio of the mix to better adapt them to the properties of the host material. It is important to stress that the selection of repair mortar composition used for renovation purposes should be carried out for each individual case taking into account the substrate characteristics as this factor influences the performance and quality of renderings.
Acknowledgements
This work was partially supported by research grant 1129/B/H03/210/38 (2010-2013) from the Polish Ministry of Science and Higher Education. The authors would like to acknowledge Dr. Julio M. del Hoyo-Meléndez for proofreading the entire manuscript.
References
[1] Weber, J., Mayr, N., Bayer, K., Hughes, D. C., Koziowski, R., Stillhammerova, M., Ullrich, D., Vyskocilova, R., 2007. Roman Cement Mortars in Europe's Architectural Heritage of the 19th Century [online], Journal of ASTM International 4(8). Available from Internet: http://www.astm.org/DIGITAL_LIBRARY/JOURNALS/JAI/PAGES/JAI100667.htm
[2] Koziowski, R., Hughes, D., Weber, J., 2010. Roman cements - key materials of the built heritage of the nineteenth century, in ""Materials, Technologies and Practice in Historic Heritage Structures". M. BoDtenaru Dan, et al. (Eds.). Springer, pp. 259-279.
[3] Scherer, G. W., 2004. Stress from crystallization of salt, Cement and Concrete Research 34, pp. 1613-1624.
[4] Espinosa, R. M., Franke, L., Deckelmann, G., 2008. Model for the mechanical stress due to the salt crystallization in porous materials, Construction and
Building Materials 22, pp. 1350-1367.
[5] Koziowski, R., Magiera, J., Weber, J., Haber, J., 1990. Decay and Conservation of Piúczów Porous Limestone. I. Lithology and Weathering, Studies in
Conservation 35(4), pp. 205-221.
[6] Benavente, D., Garcia del Cura, M. A., Fort, R., Ordonez, S., 2004. Durability estimation of porous building stones from pore structure and strength, Engineering Geology 74, pp. 113 -127.
[7] Ludovico-Marques, M., Chastre, C., 2012. Effect of salt crystallization ageing on the compressive behavior of sandstone blocks in historical buildings, Engineering Failure Analysis 26, pp. 247-257.
[8] Jaquerod, C., Chippis, Ch., Alou, F., Houst, Y. F., 1992. "Nondestructive testing of repair mortars for concrete", Proceedings of the 3rd international colloquium materials science and restoration, 1992, Esslingen, Germany, pp. 872-888.
[9] Czarnecki, L., Garbacz, A., Krystosiak, M., 2006. On the ultrasonic assessment of adhesion between polymer coating and concrete substrate, Cement and Concrete Composites 28, pp. 360-369.
[10] Wilk, D., Bratasz, L., Koziowski, R., 2010. "Reducing shrinkage cracks in Roman cement renders", 2nd Conference on Historic Mortars - HMC 2010 and RILEM TC 203-RHM final workshop. J. Válek, et al. (Eds.). RILEM Publications SARL, 1265-1272.
[11] Tsui, N., Flatt, R. J., Scherer, G. W., 2003. Crystallization damage by sodium sulphate, Journal of Cultural Heritage 4, pp. 109-115.
[12] Rodriguez-Navarro, C., Doehne, E., 1999. Salt weathering: influence of evaporation rate, supersaturation and crystallization pattern, Earth Surface Processes and Landforms 24, pp. 191-209.
[13] Rodriguez-Navarro, C., Doehne, E., Sebastian, E. 2000. How does sodium sulphate crystallize? Implications for the decay and testing of building material, Cement and Concrete Research 30, pp. 1527-1534.
[14] Vyskocilova, R., Schwarz, W., Mucha, D., Hughes, D., Koziowski, R., Weber, J., 2007. Hydration Processes in Pastes of Roman and American Natural Cements [online], Journal of ASTM International 4(8). Available from
Internet: http://www.astm.org/DIGITAL_LIBRARY/JOURNALS/JAI/PAGES/JAI100669.htm
[15] Tislova, R., Koziowska, A., Koziowski, R., Hughes, D., 2009. Porosity and specific surface area of Roman cement pastes, Cement and Concrete Research 39, pp. 950-956.
[16] Konecny, L., Naqvi, S. J., 1993. The effect of different drying techniques on the pore size distribution of blended cement mortars, Cement and Concrete Research 21, pp. 885-895.
[17] Gallé, C., 2001. Effect of drying on cement-based materials pore-structure as identified by mercury intrusion porosimetry. Comparative study between oven-, vacuum-, and freeze-drying, Cement and Concrete Research 31, pp. 1467-1477.
[18] ASTM C88-05 Standard Test Method for Soundness of Aggregates by Use of Sodium Sulfate or Magnesium Sulfate. American Society for Testing and Materials, 2005.
[19] PN-EN 1015-18:2003 Metody badañ zapraw do murów - Czçsc 18: Okreslenie wspólczynnika absorpcji wody spowodowanej podci^ganiem kapilarnym stwardniaiej zaprawy. Polish Committee for Standardization, 2003. (in Polish).
[20] Klisinska-Kopacz, A., Tislova, R., 2012. Effect of hydrophobization treatment on the hydration of repair Roman cement mortars, Construction and Building Materials 35, pp. 735-740.
[21] ASTM D7234 - 12 Standard Test Method for Pull-Off Adhesion Strength of Coatings on Concrete Using Portable Pull-Off Adhesion Testers. American Society for Testing and Materials, 2012.
[22] Klisinska-Kopacz, A., Tislova, R., Adamski, G., Koziowski, R., 2010. Pore structure of historic and repair Roman cement mortars to established their compatibility, Journal of Cultural Heritage 11, pp. 404-410.
[23] Angeli, M., Benavente, D., Bigas, J. P., Menéndez, B., Hébert, R., David, C., 2008. Modification of the porous network by salt crystallization in experimentally weathered sedimentary stones, Materials and Structures 41, pp. 1091-1108.
[24] Lanas, J., Sirera, R., Alvarez, J. I., 2006. Study of the mechanical behaviour of masonry repair lime-based mortars cured and exposed under different conditions, Cement and Concrete Research 36, pp. 961-970.
[25] Cavdar, A., Yetgin, S., 2010. Investigation of mechanical and mineralogical properties of mortars subjected to sulphate, Construction and Building Materials 24, pp. 2231-2242.
