Anales de Edificación Vol. X, X, XX-XX (2015) Received: dd-mm-aaaa ISSN: XXXX-XXXX Accepted:dd-mm-aaaa Doi: xxxxxxxx/xxxxxx

Properties of hydrated lime mortars with pozzolans

Marwa Alya & Sara Pavía b

a Trinity College Dublin (Ireland, [email protected]), b Trinity College Dublin (Ireland, [email protected])

Abstract— This is part of a wider project which aimed at finding a compatible mortar to repair monumental Portland limestone. All the lime-pozzolan mortars studied are physically compatible with the limestone as they are mechanically weaker and more permeable to liquid moisture than the rock. Most pozzolanic additions significantly increased the strength of the lime mortars however their impact on the hygric properties is lower. The pozzolans increased by up to six times the strength of the lime mortar but the RHA showed significantly lower increase proabably due to the GGBS containing a very small amount of clinkers (under 5% which is the detection limit of X-Ray Diffraction analysis) that hydrate fast providing strength. The decrease in open porosity, capillary suction and water absorption triggered by the pozzolans can be attributed to the pozzolanic hydrates leading to a greater number of very small gel pores and fewer larger pores active to liquid moisture transport. The greater decrease in open porosity, capillary suction and water absorption by the GGBS is probably due to the presence of a greater amount of hydrates. The results evidenced that lime replacement with RHA has little impact on the hygric properties; this can be attributed to the great specific surface and consequent higher water demand of the RHA increasing the amount of pores (remnants of space once filled with water). The different nature of the pozzolans affecting the pozzolanic reaction can be responsible for the variation in the lime mortar properties: the pozzolanic reaction is different in the RHA and the GGBS mortars as RHA has a greater specific surface than GGBS (which controls reactivity at the start of the pozzolanic reaction) however, the GGBS is more amorphous and contains more alumina and lime than the RHA. In addition RHA is a purely siliceous pozzolan therefore, the pozzolanic reaction results in the formation of CSH only whereas pozzolanic reaction in GGBS will produce calcium aluminum silicate hydrates (CASH) as well as CSH.

Index Terms— hydrated lime (CL90s); pozzolan; GGBS; RHA; repair mortar;

 This is part of a wider project aimed at finding a I. INTRODUCTION compatible, durable mortar to repair monumental Portland HE USE of hydrated lime (European designation: CL90s) stone, an oolitic limestone with average compressive and T as the binder in mortars involves well-known issues such flexural strengths of 54.23 and 6.23 N/mm2; 15.4% open as a slow hardening by carbonation, drying shrinkage and a porosity and 7.19% water absorption. It is widely accepted low mechanical strength however, it offers benefits such as that repair mortars for historic fabrics should be lime-based, physical and chemical compatibility with historic and and act as sacrificial materials being weaker and more traditional masonries, high workability and water retention; permeable than their substrates. and an ease of application and long-lasting plasticity that In early civilizations, pozzolans such as ceramic and allow masons to re-work mortars and reset masonry units. volcanic dust were used to enhance the properties and Furthermore, lime has some environmental credentials based durability of lime mortars and concrete. Nowadays, on a lower production energy than that of cement and the supplementary cementitious materials such as rice husk ash reabsorption of the associated CO2 emissions during hardening (RHA), ground granulated blastfurnace slag (GGBS), by carbonation. metakaolin, fly ash or silica fume are added to Portland cement and lime to improve the properties of mortars and Marwa Aly and Sara Pavía, Dep. of Civil Engineering, Trinity College concrete and reduce cement content making the materials Dublin 2, Ireland (e-mail: [email protected]) (e-mail: [email protected]). Anales de Edificación, Vol. X, X, XX-XX (2015). ISSN: XXXX-XXXX

more sustainable. lowering chloride diffusion and permeability; reducing drying Air lime hardens by carbonation; however the introduction shrinkage; increasing sulfate resistance and enhancing pozzolans alters the hardening process by producing strength. Little work has been found on the effect of GGBS on pozzolanic hydrates. The hydrates in a lime/ pozzolan matrix pure hydrated lime mortars. Işıkdağa and Bekir (2013) noted are similar to those found, upon hydration of hydraulic limes increasing strength and bond in lime/GGBS mortars. and cements; but pozzolanic hydrates form slowly and appear In the pozzolanic reaction of a GGBS/lime matrix, lime later than in cement pastes (Massazza 2002, 2007). However, Ca(OH)2 reacts with the amorphous silica in the GGBS to pozzolanic hydrates have been found after 24 h of curing in form calcium-silicate hydrates (CSH) which provide strength RHA–lime mortars, progressively increasing in size and and durability. In addition, the GGBS contains alumina amount (at 3 and 7 days) and linking to each other to form (13.85%-Table III) therefore, lime also reacts with the continuous networks after 14 days (Pavía et al. 2014). amorphous alumina in the GGBS to produce calcium This paper contributes to understanding the properties of aluminum silicate hydrates (CASH) (Osborne 1999 in Tsai et lime mortars with pozzolans by measuring the properties of al. 2014). hydrated lime mortars with GGBS and RHA (Table I). These pozzolans were selected from previous work by II. MATERIALS AND METHODS Walker and Pavía 2010, 2011 who concluded that, out of 9 Hydrated lime (CL90s), complying with EN 459-1 was pozzolans, GGBS and RHA were amongst the most reactive used as a binder. A siliceous sand similar in grading and due to their high amorphousness (non-crystalline, reactive composition to the European CEN standard sand was used as content); and that GGBS produced the highest strength aggregate. The binder/aggregate ratio for all the mortars was (together with metakaolin) followed by RHA and other high 1:3 by weight. Table 1 shows the mix proportions and water reactive-silica content pozzolans. demand for each mix. Water was added to achieve a 165±5 RHA is an agricultural waste product. In natural, mm flow diameter. The lime, pozzolans and aggregate were unprocessed rice kernels, roughly 75% is rice and bran and the dry mixed for 2 min. Water was then added and mixed for 2 remaining 25% is the husk. In countries of large rice min at low speed and finally at high speed for 1 min. The industries, the rice is par-boiled in mills which are fuelled by mortars were molded and compacted on a vibration table the husks. On combustion, the cellulose-lignin matter in the according to EN459-2. They were initially covered with damp husk burns away, leaving a porous silica skeleton which is hessian to prevent shrinkage cracking, de-molded after 3 days grinded into fine particles known as RHA (Singhania 2004). and cured for 53 days at c. 50% humidity and 20 ± 2℃ Blastfurnace slag (BS) is a by-product of the steel industry. temperature. Each property measured is the arithmetic mean of It results from the combination of iron ore with limestone flux three specimens. The Portland stone to which the repair is to and is obtained from the manufacture of pig iron in a be applied is an oolitic limestone native from the UK that blastfurnace. When BS is quenched by water it forms a glassy features heavily in historic buildings in Ireland. As material known as granulated blastfurnace slag (GBS) which aforementioned, its average compressive and flexural is later grounded (GGBS). strengths are 54.23 and 6.23 N/mm2; with 15.4% open porosity and 7.19% water absorption. A. RHA in lime mortars

RHA has been largely tested in Portland cement (PC) TABLE I matrices but hardly with lime. In the pozzolanic reaction, lime COMPOSITION AND WATER CONTENT OF MORTARS INVESTIGATED -Ca(OH)2- reacts with the amorphous silica in the RHA to Designation CL90 (%) GGBS (%) RHA (%) W/B form calcium-silicate hydrates (CSH) which provide strength and durability. According to Boateng et al. 1990 (in Billong et 100%CL 100 0 0 1.06 al. 2011) CSHs formed are CSH I and CSH II. RHA is nearly 10%GGBS 90 10 0 0.97 20%GGBS 80 20 0 0.96 pure silica (Table II and III) therefore no calcium aluminate 30%GGBS 70 30 0 0.87 hydrates occur as a result of the pozzolanic reaction. 10%RHA 90 0 10 1.00 Pavía et al. 2014 concluded that lime replacement by RHA 20%RHA 80 0 20 1.03 improves mortar workability, lowering the water/binder ratio 30%RHA 70 0 30 1.07 required to reach a specific consistency. They also concluded that RHA significantly speeds setting (a 1:3 -CL90s:RHA- A. Properties of the pozzolans mix sets 2.5 times faster than lime alone) and lowered Pozzolanic reactivity is governed by the active porosity. They also noted that increasing RHA content (amorphous) silica and alumina content and the specific significantly increases strength and elastic modulus: the surface of the pozzolan. However, according to Massaza mortars with the highest RHA content (1:3 -CL90s:RHA-) are (2007), specific surface governs reactivity at the earlier stages over 37 times stronger in compression and nearly 5 times of the pozzolanic reaction whereas later, it is the active silica stronger in flexion than the lime mortars. and alumina content that govern reactivity. The RHA in this study is a low temperature, highly B. GGBS in lime mortars siliceous and reactive ash of high specific surface area, Similarly to RHA, GGBS has been largely tested with containing little crystalline silica (cristoballite) and unburnt Portland cement (PC). It has been demonstrated that GGBS cellulose material (Pavia et al. 2014). According to these improves the general performance of PC composites by authors, its specific surface area (13.70 m2/g) is significantly

Anales de Edificación, Vol. X, X, XX-XX (2015). ISSN: XXXX-XXXX

Marwa Aly, Sara Pavía 4 superior to pozzolans such as GGBS and PFA and comparable GGBS replacement mortars have twice and five times greater to hydrated lime (16.08 m²/g), which makes lime combination strength respectively than the lime mix. easier in the early stages of the pozzolanic reaction. In The flexural strength shows a similar trend: 10 and 20% addition, despite the presence of crystalline silica (Table II) GGBS replacements increased lime mortar strength by 68 and and some carbon (maximum loss on ignition at 800° C <4%), 200% respectively and the 30%GGBS mortar reached the the RHA shows a high reactivity. highest flexural strength with a 260% increase. Replacing The GGBS has a slightly greater particle size (2-50 microns CL90s with RHA also enhances compressive strength by vs the 1-11 microns of RHA) and a much smaller specific approximately 42% (10% replacement), 153% (20% surface area than the RHA. It also contains more lime and replacement) and four times increase at 30% replacement. alumina (Table II) and is totally amorphous (Table II). The flexural strength increase for mortars with 10, 20 and 30% RHA replacement were 7, 39 and 65% respectively. TABLE II SEM analyses confirmed these results. Figure1 shows the PROPERTIES OF THE POZZOLANS (WALKER AND PAVIA 2010) typical microstructure of a hydrated-lime binder with clusters Mineralogical composition Surface Amorphousne of carbonated lime (CaCO ) forming an open structure. Pozzolan by XRD (5% detection limit) 3 area ss Figures 2 and 3 show evidence of hydration whereby a dense RHA 13.70 Total (5) Quartz , crystoballite microstructures has been created by the pozzolanic hydrates. GGBS 2.65 Mostly (4) No crystalline fraction The results are consistent with former authors reporting that GGBS produced lime pastes of highest strength, followed by TABLE III high-silica pozzolans (RHA) with a 68% reduction (Işıkdağa CHEMICAL COMPOSITION OF THE POZZOLANS (WALKER AND PAVIA 2010) and Bekir 2013; Walker and Pavía 2010, 2011). The results FE2O3 SO3 K2O Pozzolan SiO2 AL2O3 CAO also agree with Pavía et al. 2014 who reported that rising RHA RHA 93.84 1.93 0.68 0.29 - 1.38 content increased compressive and flexural strength of GGBS 34.14 13.85 39.27 0.41 2.43 0.26 hydrated lime (CL90s) mortars with CL90s:RHA (1:3) B. Physical properties of mortars mortars being over 37 times stronger in compression and The compressive (Fc) and flexural strength (Ff) were nearly 5 times stronger in flexion than lime mixes. measured according to EN1015-11. The flexural test was performed on 40x40x160 mm mortar prisms using a Zwick testing machine at rates of loading of 1mm/min. Compression 100% CL90s strength tests were carried out on the half prisms using a loading rate of 1 mm/min. The open porosity and bulk density of the mortars was tested according to RILEM recommendations. The water absorption was measured according to UNE 67-027-84. Capillary suction was measured according to EN 1925. Here, the dry samples were immersed to a depth of 3 ± 1 mm; removed and weighed at specific time intervals of 1, 3, 5, 15, 30 and 60 minutes. The coefficient of water absorption by capillarity was then calculated. C. Mortar microstructure The microstructure of the mortars was investigated using a Tescan MIRA Field Emission Scanning Electron Microscope.

The pozzolanic reaction produces hydrates which determine Fig 1. Microstructure of the hydrated-lime binder. the physical properties of the mortars (moisture movement and strength) therefore, the occurrence and nature of hydrates was 30% GGBS investigated over time.

III. RESULTS AND DISCUSSION A. Mechanical properties and microstructure As expected (Table 4) pozzolan replacement enhances lime mortar strength which increased with increasing replacement level. GGBS mortars achieved higher strength than RHA mortars at all replacement levels. This was expected as GGBS contains a very small amount of clinkers (under 5% which is the detection limit of X-Ray Diffraction analysis), similar to those in PC, that hydrate fast providing strength, being therefore a latent hydraulic binder rather than a pozzolan. The 30%GGBS replacement mortar has a compressive strength 6 times greater than the lime mortar. The 10 and 20% Fig 2. Microstructure of the hydrated-lime binder including 30% GGBS.

Anales de Edificación, Vol. X, X, XX-XX (2015). ISSN: XXXX-XXXX

hydrates such as CSHI and II results in a lime mortar with a greater number of very small gel pores (under 550 Å = 0.05 30% RHA microns-µm) (Lea 2003) and fewer larger pores active to liquid moisture transport. Benavente (2011) identifies that liquid water is transported by capillary forces in pores greater than 0.1 μm with high water absorption rates in the size interval between 1 μm and 1 mm. The greater decrease in open porosity, capillary suction and water absorption by the GGBS is probably due to the presence of a greater amount of hydrates (with nanopores out of the range of liquid moisture transfer) blocking capillary pores.

TABLE IV MECHANICAL AND HYGRIC PROPERTIES OF MORTARS AT 56 DAYS. FC- COMPRESSIVE STRENGTH; FF-FLEXURAL STRENGTH; P-OPEN POROSITY; W- WATER ABSORPTION; S-CAPILLARY SUCTION. W S F F P Mortar C F (MPa) (MPa) (%) (%) (kg/m2.min0.5) Fig 3. Microstructure of the hydrated-lime binder including 30% RHA. 100%CL 2.48 0.41 32.36 20.40 2.49 B. Hygric properties 10%GGBS 4.94 0.69 32.38 20.52 2.25 20%GGBS 12.33 1.27 31.96 19.48 1.89 Pozzolan replacement did not significantly affect the 30%GGBS 14.87 1.48 26.83 15.86 1.04 properties of the lime mortars (except for the 30% GGBS 10%RHA 3.53 0.44 31.49 18.98 2.45 replacement that lowers porosity, water absorption and suction 20%RHA 6.29 0.57 30.74 18.69 2.37 by 17%, 22% and 58% respectively). 30%RHA 10.18 0.68 30.32 18.01 2.14 The 26-32% open porosity of all lime-pozzolan mortars is comparable to the 18-40% values documented by Moropoulou et al. (2005) for ancient lime mortars in the Mediterranean IV. CONCLUSION Basin and compatible with the 15.4% open porosity of the All the lime-pozzolan mortars studied are physically Portland limestone. The hygric properties of the control mortar compatible with the Portland limestone as they are are consistent with previous authors (Cosgrove and Pavía mechanically weaker and more permeable to liquid moisture 2009) who reported 34% porosity, 19% water absorption and than the rock. 2.71 kg/m2.min0.5 capillary suction for CL90s mortars made Most pozzolanic additions significantly increased the with standard sand at B/A ratio of 1:3 by weight. strength of the lime mortars however their impact on the 30% GGBS replacement lowered the lime mortar porosity hygric properties is much smaller. The pozzolans increased by by approximately 17%. This agrees with Griffin 2004, who up to six times the strength of the lime mortar but the RHA reported that addition of GGBS to hydrated lime in grouts showed a significantly lower strength increase. This was decreased porosity by c.25%. 20% GGBS replacement slightly expected as GGBS contains a very small amount of clinkers decreases the porosity by about 2% and no reduction was (under 5% which is the detection limit of X-Ray Diffraction observed at 10 % GGBS replacement. analysis), similar to those in PC, that hydrate fast providing The results evidenced that lime replacement with RHA has strength. little impact on the porosity; this is probably due to the much The results evidenced that lime replacement with RHA has greater specific surface and consequent higher water demand little impact on the hygric properties; this can be attributed to of the RHA increasing the amount of pores (it is well known the great specific surface and consequent higher water demand in concrete technology that some pores are remnants of space of the RHA increasing the amount of pores (remnants of space once filled with water (Walker and Pavía 2010, 2011). once filled with water). According to Papayianni and Stefanidou (2006) water/binder The decrease in open porosity, capillary suction and water ratio is the most important factor influencing porosity in lime absorption in the pozzolan mortars can be attributed to the mortars. pozzolanic hydrates leading to a greater number of very small A similar trend was observed for the water absorption, as gel pores and fewer larger pores active to liquid moisture this property strongly relates to the open porosity. According transport. The greater decrease in open porosity, capillary to the results, 6 to 12% water absorption decrease was suction and water absorption by the GGBS is probably due to observed for the RHA mixes and over 22% decrease for the the presence of a greater amount of hydrates. 30%GGBS mortar (with the lowest porosity). The different nature of the pozzolans affecting the The capillary suction results are consistent with the above. pozzolanic reaction can be responsible for the variation in the A 20-30% GGBS replacement lowered the capillary suction of lime mortar properties: the pozzolanic reaction is different in the lime mortar (by 24 and 58% respectively) but the RHA the RHA and the GGBS mortars as RHA has a greater specific replacement didn’t significantly impact suction (the highest surface than GGBS (which controls reactivity at the start of RHA replacement only lowers suction by 14%). the pozzolanic reaction) however, the GGBS is more A decrease in open porosity, capillary suction and water amorphous and contains more alumina and lime than the absorption was expected as the formation of pozzolanic

Anales de Edificación, Vol. X, X, XX-XX (2015). ISSN: XXXX-XXXX

Marwa Aly, Sara Pavía 6

RHA. In addition RHA is a purely siliceous pozzolan Singhania N. P. (2004). “Rice Husk therefore, the pozzolanic reaction results in the formation of Ash.” The Institute of CSH only whereas pozzolanic reaction in GGBS will produce Journal Article, Concrete Technology (Singhania, 2004) calcium aluminum silicate hydrates (CASH) as well as CSH. 1 Authors Newsletter, The Engineering Council (UK), Issue 55. ACKNOWLEDGMENT The authors wish to thank the Irish Research Council and Billong, N., Melo, U. C., Kamseu, the Office of Public Works for funding the project. We also E., Kinuthia, J. M., & Njopwouo, D. (2011). thank John Cahill, Conservation Architect of the Office of Improving hydraulic Public Works, for facilitating investigation and site work in properties of lime–rice (Billong et al., Journal Article, husk ash (RHA) binders the project. The testing was carried out in the Department of 2011) Civil, Engineering, Trinity College Dublin. The authors thank 3-7 Authors with metakaolin (MK). Construction and Chief Technician Dr. Kevin Ryan, for his help with the Building laboratory work and our technicians Dr. Michael Grimes, Mr. Materials, 25(4), 2157- Eoin Dunne and Mr. Dave McAuley for their assistance with 2161 testing. Işıkdağ, B. and İ. B. Topçu (2013). The effect of ground REFERENCES granulated blast-furnace Journal Article, slag on properties of (Işıkdağa and Bekir In Reference List Horasan mortar. Reference Type In-Text Citation 2 Authors 2013) Construction and Building Materials 40: Massazza F (2002) Properties and 448-454. applications of natural pozzolans. In: Bensted J, Tsai, C. J., Huang, R., Lin, W. T., & Book: 3-6 Barnes JP (eds) Structure (Massazza, 2002) Wang, H. N. (2014). Authors and performance of Mechanical and cements. Spon Press, cementitious London Journal Article, characteristics of ground (Tsai et al., 2014) 3-7 Authors granulated blast furnace Massazza F (2007) Pozzolana and slag and basic oxygen pozzolanic cements. In: furnace slag blended Hewlett PC (ed) Lea’s mortar. Materials & Book: 3 to 6 chemistry of cement and (Massazza, 2007) Design, 60, 267-273. Authors concrete, 4th edn. Elsevier, UK, pp 471– EN 459-1 (2010). Building lime. 602. Part 1: Definitions, specifications and Pavía, S., Walker, R., Veale, P. & Building conforming criteria. (EN 459-1) Wood, A. (2014). Impact standard European Committee for of the Properties and Standardisation CEN, Journal Article, Reactivity of Rice Husk Brussels. (Pavia et al., 2014) 2-6 Authors Ash on Lime Mortar Properties. Journal of EN 459-2 (2001). Building lime. Materials in Civil Building Test methods. European (EN 459-2) Engineering, 26(9). standard Committee for Standardisation CEN. Walker, R., and Pavía, S. (2010). “Behaviour and EN 1015-11 (1999). Methods of test Properties of Lime- for mortar for masonry. Building Pozzolan Pastes.” Proc., Determination of flexural (EN 1015-11) Journal Article, (Walker & Pavía, standard and compressive strength 2 Authors 8th Int Masonry 2010) Conference, International of hardened mortar. Masonry Society, Dresden, 353-362. RILEM (1980). Recommended tests to measure the Walker, R. and S. Pavìa, Physical deterioration of stone properties and reactivity Technical and assess the (RILEM 1980) Journal Article, of pozzolans, and their. (Walker & Pavía, recommendation effectiveness of 2 Authors Materials and Structures, 2010) treatment methods. 2011. 44: p. 1139–1150. Materials and Structures 13: 175– 253.

Anales de Edificación, Vol. X, X, XX-XX (2015). ISSN: XXXX-XXXX

UNE 67-027-84. Determinacion de la Building absorcion de agua. (UNE 67-027-84) standard Ladrillos. EN 1925 (1999). Natural stone test methods. Determination Building of water absorption (EN 1925) standard coefficient by capillarity. P Moropoulou, A., A. Bakolas, and S. Anagnostopoulou, Composite materials in Journal Article, (Moropoulou et al. ancient structures. 3-7 Authors 2005) Cement and Concrete Composites, 2005(27): p. 295–300. Cosgrove, K., and S. Pavía. Mechanical and Fluid Transfer Properties of Conference some Lime and Portland (Cosgrove & Pavía paper Cement Mortars. 2009) PROHITEC 2009 Int.Conf.

Griffin, I. (2004). Pozzolanas as Additives for Grouts. Journal Article, STUDIES IN (Griffin 2004) 1 Author CONSERVATION 49: 23-34.

I. Papayianni, M. Stefanidou. (2006) Strength-porosity relationships in Lime- Journal Article, Pozzzolan mortars, (Papayianni, & 2 Authors Construction and Stefanidou 2006) Building Materials,Vol. 20 Issue 9

Lea’s chemistry of cement and concrete ed by P.C. Book: 3 to 6 Hewlett 2003. Lea’s (Lea 2003) Authors chemistry of cement and concrete, 4th edn. Elsevier, UK,

Benavente D. (2011) Why Pore Size Is Important in the Deterioration of Porous Journal Article, Stones Used in the Built (Benavente 2011) 1 Author Heritage. Revista de la sociedad española de mineralogía; 2011:15.

Anales de Edificación, Vol. X, X, XX-XX (2015). ISSN: XXXX-XXXX

Anales de Edificación, Vol. X, X, XX-XX (2015). ISSN: XXXX-XXXX