MECHANICAL PROPERTIES & DURABILITY OF CEMENT-STABILIZED EARTH MORTARS FOR APPLICATION IN PREHISTORIC MONUMENTS

Eleni-Eva Toumbakari*

Directorate for Restoration of Ancient Monuments, Hellenic Ministry of Culture, 12, Karytsi Sq., 10561 Athens, Greece

Maria Kaipanou, Aphroditi Ntziouni and Vassileia Kasselouri-Rigopoulou

School of Chemical Engineering, National Technical University of Athens 9, Iroon Polytechniou str., 15773 Athens, Greece

Keywords: Earth, Mortar, Conservation, Prehistory, Durability ABSTRACT The present work deals with the properties, mainly physical and mechanical, of repair mortars destined for the repair and strengthening of prehistoric masonry. The paper focuses on a specific prehistoric building type, the tholos tomb. Mixtures composed of Portland cement in limited quantities, clay and calcareous sand are examined. Aim of the mix design is to produce repair materials with Portland cement content as low as possible, to avoid mainly durability failures. Those materials should at the same time be able to ensure effective repair and strengthening, which is required in earthquake-prone areas such as Greece. The follow up of the evolution of some key parameters with time, such as compressive and tensile strength and apparent porosity, has shown that the cement-stabilized mortars can develop a great variety of properties, some of which can satisfy historic masonry requirements. Moreover, the study of the microstructure, using Scanning Electron Microscopy and X-Ray Diffraction, revealed the development of pozzolanic reactions, which is a key-factor for durability. However, a minimum percentage of cement is required, in order to improve the setting time and to ensure durability against sulphate attack as well. Cement-stabilized earth compositions permit the reduction of the Portland cement content and are able to produce a variety of mechanical and physical properties, sufficient for the repair and strengthening of prehistoric masonry structures as well as for use in archaeological sites. INTRODUCTION

Earth-based binders were often used in the construction of prehistoric buildings in Greece. The function of those binders was either bedding mortars or plasters. The paper focuses on a specific prehistoric building type, the tholos tomb, where such binders were occasionally used. A tholos tomb consists of a circular burial chamber, called thalamos, initially roofed by a corbelled vault (Figure 1) and approached by an entrance passage, the dromos, which narrows abruptly at the doorway, the stomion. The thalamos is built of stone courses. The “prehistoric” way of assembling the stone blocks consists in the direct application of a stone on top of other, without the mediation of a mortar, as in Roman, Byzantine and later masonry assemblages. In the case, however, of irregular masonry, this technique results in the creation of, smaller or bigger, voids between the stone blocks which are often found filled with earth (Figure 2a). The fact that this earth fill was not systematically found during the excavations, led some researchers to the hypothesis that the tholos tombs were constructed solely of dry-stack masonries. The size, however, of many tholos tombs as well as stability requirements during construction call in favor of the use of this earth fill between the building stones as well. This assumption could be applicable to the majority of the tholos tombs, mainly those made of roughly carved stones, and it still remains an open subject.

Figure 1: Example of a Tholos Tomb: interior of Tholos IV in Thorikos, Attica

Focusing on the structural pathology of the masonries envountered at the interior of many tholos tombs, it is systematically observed that the long-term exposure to environmental strains resulted in the creation of voids at the interstices between the stone blocks and layers (Figure 2b). Deterioration is then rapidly initiated, and sometimes leads to severe structural problems. A very common structural problem, in case of material loss, is the bending failure of the building stones, which were initially meant to lye one over the other (Figure 2b). If deformation proceeds, then local instabilities could take place, and eventually lead later to collapses. In this case, one of the techniques that could be used for the structural restoration of the tholos tomb masonries is the application of a mortar at the interstices between the stones, in replacement to the original earth fill (if it was) or simply in order to fill the voids created by material deterioration. This measure permits the increase of the frictional internal forces of the system and the overall Modulus of Elasticity.

(a)

(b)

Figure 2: (a) Example of irregular masonry (tholos tomb of Acharnae, Attica), (b) severe stone deterioration and loss of material between the stones (Tholos IV in Thorikos, Attica)

Consequently, a more uniform stress distribution is expected to be restored, reducing the risk of stress concentrations, which lead to unwanted bending or even out-of-plane deformations, such as the ones shown in Figure 2b. On the basis of the aforementioned considerations, it is decided to study and develop suitable earth mortars for application in prehistoric masonries.

EXPERIMENTAL

Selection and characterization of the materials It is well-known that one of the requirements for the design of mortars for application in historic masonries is the study of the original materials in order to produce mortars, physicochemically and mechanically compatible to the original fabric [1]. Earth mortars are, however, vulnerable to environmental conditions and require protection to achieve durability. An alternative could be the design of stabilized earth mortars [2-4]. These are earth-based mortars, in which a part of the binder is typically replaced by cement or lime. To produce the mortar, an earth binder industrially produced in Attica was selected. Its grain size distribution is shown in Figure 3 [5]. In Figure 4 the morphology and the composition of the clay binder are also presented. The material mainly consists of albite and silica as it is also confirmed by the XRD pattern (Figure 5).

Earth grain-size distribution

0

10

20

30

40

50

60

70

80

90

100

0,001 0,01 0,1 1 10 100

Diameter [mm]

Passing [%]

Figure 3: Grain-size distribution of earth binder

Figure 4: SEM image (x3000) and EDX Spectrum of earth binder

0

100

200

300

400

500

600

700

800

900

1000

0 10 20 30 40 50 60

2θo

cps

Figure 5: XRD pattern of earth binder

On the basis of the grain-size distribution, it is concluded that the earth is composed of fine-grain sand (2.9%), silt (79.8%) and clay (17.3%). The corresponding Atterberg limits are presented in Table 1.

Table 1: Atterberg Limits Liquid Limit LL 38.2 % Plasticity Limit PL 22.2 % Plasticity Index PI 16.0%

Selection of the mortar compositions In the present study, two stabilized earth mortar compositions were studied. The binder to aggregate ratio was kept constant and equal to 1:1 per weight. Calcareous sand of Attica was used as aggregate, whereas the binder consisted of silt and clay, to which cement type CEM II 32.5N in proportions of 12.5 % and 25%-wt was added. The water content was defined on the basis of flow-table tests. A requirement for a 16-17 cm mortar expansion was set. The compositions are presented in Table 2.

Table 2: Composition of the earth-based mortars Composition Clay-silt Cement Sand

A 25% 25% 50% B 37.5% 12.5% 50%

Curing conditions Mortar preparation took place with a standard Hobart-type mixer. The specimens were prepared and cured as follows: first, the specimens were allowed to stay in the moulds for 7 days before

• Quartz ¤ Albite * Muscovite + Montmorillonite

•

¤

* *

+ +

•

demoulding, because of their slow strength increase. They were stored inside the moulds in the wet chamber for 7 days. Then, they were carefully removed from the moulds, and were stored in a laboratory closet under environmental conditions (22°C at approximately 60-65% R.H.) until the day of testing. Specimen destined for mechanical tests had dimensions 40x40x160mm whereas those used for durability tests were measuring 25x25x285mm following Norm ASTM C 490. Fragments of the mortars, obtained after the mechanical tests, were immediately immersed once in acetone and twice in diethyl ether and then let to dry under vacuum overnight, in order to stop hydration. The fragments were subsequently stored in air-tight vessels to avoid carbonation until the day of testing.

Experimental procedures A very important part of the study was the assessment of the mortars’ hygric properties and durability. The mortars’ hygric behaviour was studied through the determination of the water absorption and drying capacity of the mixtures at the age of 28 days. Durability was studied through constant immersion of the mortar specimen in deionized water and in sodium sulphate solution. At specific intervals, the specimens were removed from the water or the solution and their length and weight were measured. Then, the specimens were put back to water or the solution until the next measurement. The mortars were also subjected to cycles of wetting and drying. Moreover, the mortar shrinkage under exposure in laboratory conditions (22°C and 60%R.H.) was equally monitored. To measure open porosity at a certain age, the specimens were first oven-dried at 60°C until constant weight. After having cooled off, they were immersed in water until weight stabilisation. The saturated sample was then hung in a weight balance and immersed in water for volume determination. Concerning the microstructural study, XRD analysis was performed with a Siemens D5000 Diffractometer(Cu Ka, k = 1.5406 A˚), whereas Scanning Electron Microscopy coupled with EDX microanalysis was performed for the observation of the mortar fracture surfaces with an Jeol6380LV Analyzer.

RESULTS & DISCUSSION

Mechanical properties, drying shrinkage & open porosity It is well-known that earth binders present important volume changes as a function of the environmental conditions. The length change of compositions A and B was measured as soon as the samples were removed from their moulds. The results are shown in Figure 8.

Drying shrinkage

90

91

92

93

94

95

96

97

98

99

100

0 7 14 21 28 35 42 49 56 63 70 77 84 91 98 105 112 119 126

age [days]

length change [%]

comp A

comp B

Figure 6: Evolution of the mortar drying shrinkage

Table 3: Physicomechanical properties of composition A Age [days]

Flexural strength [MPa]

Compressive strength [MPa]

Open porosity [%]

7 1.87 6.97 29.73 28 2.87 11.58 31.14 90 4.13 19.05 27.19 180 4.00 19.57 29.38

Table 4: Physicomechanical properties of composition B

Age [days]

Flexural strength [MPa]

Compressive strength [MPa]

Open porosity [%]

7 0.7 1.57 36.86 28 1.1 3.43 36.63 90 1.5 5.52 36.41 180 1.17 5.45 35.48

Both compositions present the same tendency; they differ however in the importance of the measured shrinkage. Shrinkage of the mortars is observed until practically the first month (35 days), after which they seem to stabilize in the given laboratory conditions. The shrinkage values are quite limited: 2% and 3.5% for compositions A and B respectively. The evolution of the mortar mechanical properties of the mortar compositions A and B are presented in Tables 3 and 4 and in Figures 6 and 7. As expected, the flexural and compressive strength achieved by composition A are higher than those of composition B, due to the higher cement content. It is interesting to observe that an increase in the compressive as well as the flexural strengths of the mortars takes place until the age of 90 days. After this age, the compressive strength of both compositions seems to remain constant, whereas a small decrease of the flexural strength of composition B is observed. Porosity is an important factor affecting durability. As expected, composition B, with lower cement content, exhibits higher open porosity values in comparison to composition A. The open porosity of both compositions remains however quite high: a value of approximately 29% was measured for composition A and 36% for composition B. However, some loss of material during testing cannot

Flexural strength

0

0,5

1

1,5

2

2,5

3

3,5

4

4,5

5

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180

age [days]

f c [MPa]

comp A

comp B

Figure 7: Evolution of the mortars flexural strength

Compressive strength

0

5

10

15

20

25

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180

age [days]

f c [MPa]

comp A

comp B

Figure 8: Evolution of the mortars compressive strength

be excluded. This loss of material could slightly affect the final porosity values. In any case, the open porosity is sufficiently high, so as too be compatible with the majority of stone types used for the construction of most Tholos Tombs in Greece.

Microstructural Study X –Ray Difractometry and Scanning Electron Microscopy were used for the study of the hydration products of both mortar compositions A and B at the age of 28 and 90 days respectively ( samples A28, A90, B28 and B90). X-RayDiffraction measurements

0

100

200

300

400

500

600

700

800

0 10 20 30 40 50 60

2θ

cps

A28

A90

B28

B90

Figure 9: X-ray Diffraction patterns of the earth mortars of both compositions at ages 28 and 90

days

*Calcite

• Quartz † Tobermorite 9Å ¤ Ca(OH)2

*

•

††

* ¤

In general, the hydrated products tobermorite 9Å, truscottite and Ca(OH)2 as well as CaCO3 either as the main sand constituent or from a partial carbonation of the Ca(OH) 2 released, are presented. The peaks concerning SiO2 (one of the main constituent of the earth used) show a further reduction in the case of sample B90 in comparison with the sample B28. Taking into account that the series B contains the half amount of cement, which leads to less Ca(OH)2, it confirms a pozzolanic reaction between binder and Ca(OH)2.

Scanning Electron Microscopy (SEM) measurements Calcium silicate hydrates are observed either into the pores or on the surface of clay grains. Especially in the case of B series (Figures 9 and 10), even though less cement has been added (12,5 wt%) in comparison with composition A, meaning less Ca(OH)2 release, a satisfied reaction of clay grains which leads to calcium silicate hydrates, has been observed.

Figure 10: SEM image (x6000) and EDX Spectrum of sample A28

Figure 11: SEM image (x6000) and EDX Spectrum of sample A90

Figure 12: SEM image (x6000) and EDX Spectrum of sample B28

Figure 13: SEM image (x6000) and EDX Spectrum of sample B90

Durability in water and in sulphate solution Very often, the vaults of the Tholos Tombs have collapsed and the interior is exposed to rain water. When drainage is insufficient, water stagnates at the base of the monument until it dries. There exist also cases, where the Tholos Tomb is periodically inundated by underground water. Consequently, durability of the mortar, when immersed in water, is also very important. The evolution of the length change of the two compositions A and B, when constantly immersed in deionised water, is presented in Figure 14. The two compositions exhibit a similar behaviour. Indeed, after an initial shrinkage attributable to the air-curing already described, the compositions exhibit a slight length increase as soon as they are put in the water. Practically some hours after immersion, the specimen is filled with water and then it exhibits remarkable stability.

Length change under constant immersion in deionised water

90

91

92

93

94

95

96

97

98

99

100

0 7 14 21 28 35 42 49 56 63 70 77 84 91 98 105 112 119 126

age [days]

length change [%]

comp A

comp B

Figure 14: Evolution of the mortar length change after constant immersion in deionised water

(up to 28 days: standard curing in air) The evolution of the length change of the two compositions A and B, when constantly immersed in sulphate solution, is presented in Figure 15. Composition B failed very quickly. On the contrary, composition A, with a higher cement content, exhibited a length increase, and despite the presence of small hairline cracks, it did not fail. It has to be noted here that the acidity of the solution was quickly attenuated and a PH around 11 was restored. The solution was often renewed, nevertheless it still became alkaline. Further research is required, in order to check the mortar durability when the solution is constantly acid.

Length change under constant immersion in sodium sulphate

90

92

94

96

98

100

102

104

106

108

110

0 7 14 21 28 35 42 49 56 63 70 77 84 91 98 105 112 119 126

age [days]

length change [%]

comp A

comp B

Figure 15: Evolution of the mortar length change after constant immersion in sodium sulphate

(up to 28 days: standard curing in air)

CONCLUSIONS

In order to develop earth-based mortars for the repair and strengthening of Tholos Tombs in Greece, a research programme was carried out. The main results of the research could be summarized as follows:

1. Earth-cement mortars can exhibit a great variety of mechanical properties, which makes them suitable for various applications.

2. The open porosity of the compositions is compatible to the majority of stone types used for the construction of most Tholos Tombs.

3. The first tests showed that the studied compositions present limited shrinkage and very good durability in deionised water.

4. When exposed to sulphate solutions, the mortar with the lower cement content failed very quickly. In order to increase resistance to sulphate, for the given binder to sand ratio, the increase of the cement content is necessary.

5. XRD and SEM measurements indicate the existence of pozzolanic reaction between the binder and the Ca(OH)2 released during the cement hydration process.

ACKNOWLEDGMENTS

The physico-mechanical tests were performed at the EKET (Hellenic Cement Research Centre) Laboratory. Mrs Z.Tsimpouki, EKET General Director, Dr V.Kaloidas, EKET Technical Director and Mr C.Alafouzos, Resp. of the Laboratory for Physico-mechanical Tests are gratefully acknowledged.

REFERENCES

[1] Rilem TC 203-RHM, (2009), “Repair mortars for historic masonry. Testing of hardened mortars, a

process of questioning and interpreting”, in Materials and Structures, Vol. 42, No 7, pp. 853-865. [2] Bei G., (1996), “Raw earth – an ancient & modern building material”, M.Sc.Diss. (Supervisors: D.Van

Gemert, K.Van Balen, I.Papayianni), Kath.Univ.of Leuven – R.Lemaire Centre of Conservation, p. 110 + Appendices.

[3] Zinn W., (2005), “Cement modified earthen mortar- an investigation of soil-cement performance characteristics at three Southwestern National Monuments”, M.Sc.Diss. (Supervisor: F.Matero), Univ.of Pennsylvania, p. 300.

[4] Goodbey R.J., Thomson M.L., (2008), “Compressed earth block: achieving building code requirements with lime stabilization”, in J. of the Masonry Society, Vol. 26, No. 2, pp. 9-18

[5] ASTM D 2487-06, (2006), “Standard test method for classification of soils for engineering purposes”, ASTM International, Pennsylvania.