Construcon and Bung Mater 64 (2015) 125-135 Contents lists available at ScienceDirect Construction and Building Materials ELSEVIER journal homepage: www-elsevier.com/locate/conbuildmat Bond-strength performance of hydraulic lime and natural cement mortared sandstone masonry S. Barr®, WJ, McCarter”, B. Suryanto” “engineering Design, Development Series Flitk Coun. Abo Hous, akin R22 Solan UK Seo Energy, CescenceInaratre en cet, eat Watt river. Ebr aS Sand + Fexural and compressive stress-strain response of NHL and nat {Morea stitiness correlates with increasing hydraulic of binder {New data presented on the bond strength of martared sandstone masonry unit. ‘intluence of pre-wetting time on bond stengih of mortared sandstone units presente, {sandstone masonry bond strength increase: linearly with increasing hycralcty of binder, al cement mortar presente. ARTICLE INFO ABSTRACT ‘ae an ‘Availabe online 19 March 2015 Flexural bond sengih is an important pevformance characteristic of masonty swucures yet there is no auidance for lime-mertared stonework in design codes of practice. This study investigates the bond ‘Nrength of natural hydraulic lime (NHL) and natural cement mortared sandstone masonry. To this fend, the flexural Bond strength of masonry couplets, built with mortars of three hydraulic strengths and one natural cement and having a water-costent adjusted to achieve a simular consistency. was mee- Sured with the bood wrench test. Practical mortar compositions and natural curing conditions were used within the experimental programme. Bond strength was found tobe dtety related to binder hydraul- ‘iy and sandstone pre-wetting time ~ 4 positive elfect in the ease ofthe former anda negative influence in the case of the ler, Pre-weting Kime, however, hada greater snfhience onthe feeby hyeraule Hime binder (NHL 2) than on the moderately (NHL 3.5} and eminently hydraulic (NHL 5) lime binders. The ess presented will assist in improving our Knowledge of lime mortared sandstone masonry and in Rewordr the development of design guidance (© 2015 Elsevier Utd All ight reserve. 1. Introduction Lime mortared brickwork and stonework has been used in masonry construction since ancient times. However, the use of lime mortared masonry has been largely displaced since the advent of stronger and faster setting modem Portland cement in ‘the late 19th century. It became evident in the late 20th century ‘that inappropriate use of cement mortars lead to accelerated masonry deterioration [1 which did not occur with lime-mortared masonry due to lime mortars greater breathability (2) tn addition to its breathability, a lime mortats ability to accommodate move- ment and its aesthetic appeal has, in recent years, driven a ~FCrrsponcing uth. Te: 1481214512308 fas 0441314514817 mal adress mech ac (W) Meare) resurgence in its use in masonty, particularly in sandstone mason- Fy construction and conservation projects, Hydraulic ime mortars, such as Natural Hydraulic Lime (NHL) and natural cement (NC) mortars can set underwater and gain strength by both hydration and carbonation reactions, unlike air- lime mortars which gain strength purely by carbonation. Hydraulic lime mortars are both faster setting and stronger than air-lime mortars but have greater permeability and teduced sti ress in comparison to Portland cement mortars [2|, Despite the advantages of hydraulic lime mortars over cement mortals, theit Use is inhibited by 2 lack of published design guidance and perfor- ‘mance data. Ths lack of data also prevents accurate assessments of the considerable quantity of existing masonry structures built from hydraulic ime mortared natural stone. The flexural bond strength of masonry is a particularly important performance characteristicWhich plays a significant role inthe ability of a masonry structure to resist lateral or eccentric loading. No specific mention of lime mortars is made in Euracode 6 [2) Eurocode 6 (ECS) uses performance based limit state design with mortars being designated according to compressive strength in a standard 1:3 binder: aggregate mix ratio by mass, Within ECB other strength characteristics, such as masonry shear strength and ‘masonry flexural strength, are derived from the mortar compres- sive strength, Lime mortars have much lower compressive strength than cement mortar and so for EC6 design, all other lime mortar ‘masomy strength characteristics are automatically designated within or below the lowest category. According to ECG, the mason- fy flexural strength parallel tothe bed joints fora standard mortar with compressive strength under $Nfmm* and natural stone ‘masonry is 0.05 NJmm* ~ a very low value. The UK national annex to EC6 [4] includes no masonry flexural strength data for natural stone masonry or for mortar under 2 Nimm® compressive strength. There has been increasing academic interest in the flexural strength of lime-mortared masonry, likely driven by an increasing awareness of the benefits of lime mortared masonry and general paucity of bond strength data, Work has, in the main, focussed ‘on modern clay bricks of various types. For example, Zhou et al [5] tested clay bricks (perforated and unperforated) of various absorptivity with hydraulic limes mortars of various hydraulicties and mix ratios for curing periods up to $1 days. Not all configura- tions were tested and experiments focussed on a NHL 3.5 mortar in 21:22 mix rato by volume (1:6.62 by mass) using dy clay bricks With initial rate of absorption ranging from 0.1 to 2.4 kg/(m* min). For a NHL3.S mortar in 4 1:225 mix ratio by volume, mean values fof masonry flexural strength (based on the bond-wernch test) were found to range from 0.08 Nimm! for the highest suction brick to (0.49 Njmm? for a medium suction brick, The highest value of masonry flexural strength was found to be 0.63 Nimm? for a NHL 5 mortar in a 1:2.25 mix ratio by volume with a medium suction brick, Pavia and Hanley [5] also tested clay bricks which were pre-wetted to control suction using lime mortars of various hhyeraulicity and flow in a 1:2.5 mix ratio (by mass) for a curing period of 28-days. For a NHL 3.5 mortar mean values of masonry flexural strength were found to range from 0.20N/mm? for a low-flow mortar to 0.61 Nim? for a high-flow mortar. Mean val- tues of masonry lexural strength for NHL2 and NHL mortars ran- ged between these values and generally increased with greater ‘mortar hydraulicty and flow. Costigan and Pavia [7] tested dy, medium suction, Frogged clay bricks with hydraulic lime mortars of vatying hydraulicity in a 1:3 mix ratio (by mass) fora range of curing periods. For a curing period of 28-days, mean values of, masonry flexural strength were approximately 0.11 Nimm# for NHL 2 mortar, 0.16 N/mm? for NHL. 35 and 0.15 Njmm? for NHL 5 mortar. For a curing period of &-months, mean values of masonry flexural strength had increased ané were, approximately, 0.19 Nj rmm® for NHL2 mortar, 0.40 Nim for NHL 3.5 and 0.37 Njmm? for NHL 5 mortar, Lawrence et al [8] identiied critical brick-surface pore sizes that govern bond strength. It was found that calcium silicate crys- tals can penetrate pores under 1 pm whereas calcium hydroxide cystas can only penetrate pore sizes above 1 jum. It was conclud- ed that bond strength of hydraulic ime mortars would improve with greater proportion of brick-surface pore sizes under 1 ym. Other related studies on mortars include those Hendrickx et all [9] investigating the early water transport between two mortars of different water retention ~ a lime-mortar and a cement-mortar ~ and two bricks of different absorption rates ~ an extruded clay brick and a moulded clay brick. Both mortar water retention and block absorption rate influence the amount of residual water femaining in the mortar and it was concluded that the effect of ‘mortar water retention on water transport is greater than the brick absorption rate. Aggregate texture, size and grading all influence the workability, compressive and flexural strength of mortar |10~ 13) which wall in turn, effect bond strength. It should be noted that previous studies have used lime mortars with lay bricks asthe block material, In practice, however, there is a much greater need for natural stone to be paired with hydraulic lime mortar due to stone being generally more susceptible to dete- rloration caused by cement mortar. This study aims to characterise the flexural bond strength of hydraulic lime mortared sandstone blocks and to determine the correlation between masonry flexural strength, mortar bed-joint strength and block absorption (pre-wet- ting time), Prompt natural cement mortar is also investigated; in addition, mortar mix ratios commonly used in practice (batched by volume) are employed together with natural curing conditions that would be experienced on site. Regarding pre-wetting, studies fn clay bricks have shown that pre-welting can have either a posi- tive of negative effect on the interfacial bond [1,15], therefore it was also the intention of this study to clarify this matter for sand- stone blocks, 2. Experimental programme Cull stone a ane-ganed ey sandstone fom the Culaloe quay i ile (scotland, was supplied in brcksized dimensions ve. 215 « 1025 » 65 mm 16) total absorption ~ 5%. The coefficient of water absorption due to capillary action a5 fread SEN 7721-20111 | ena a mandory tert for supplier fo ep a permed 06% cine sc aeetive tear the set of NC were used trout, [ISTO] A wel-gaded buleig sané 2 man maximal pasice ste) was wed Fefecing common ste practice 2.2. Moc proporns on inl fo Mortars were pe-bagged in 412 Laeaggregate mi sai by volume (at nas) a5 commonly wed In proctce The consistency of the mortar mix Was ered by mensremen of the inal flow sn scoréance with BS ENTOYS 531999 20 Teemu acequate werhabiy, and trepleste common ste pact, an intl fow of approximately 170cnm was speciied The water demand te aieve smlar consistency eeceased wh lnetessng ime gage: 25 2 conse uence, the wate-contet deceased by ss 10% lor the prescribed consistency ‘mixes used within the experimental programme ae presente i 12912 2.2, Woe berpton of Cull andtone 7721-2011 17 an based on the Fess fom 5k (atonal) sent ane {edb eneesing tebe faces fac lok to depth ofS ray of water orto bonding. ive snimession tines were considered: Oust lek) ub Immersion ands mn immesson, 24 Block bonding. workmanship and cing aso coupes, as per 55 EN 1052-52005 |2) of bonded blocks were pre paced foreach af the dry I min and 15min pre-wetung ates and fr eath of rable a ineclgil and pic dats for he ingested G5 alte) ° ° ° 16 = ele ” 3s % 30 Ca(OH) ee time') 58 2s 2 2 Density es) 50650 70 treobo 5 Bay ea /Comruction and Bling Materia 842015) 125-135 rable Mortar batch prepeton init! ow (binders) Grade Deas ag) Ma gies) SST) Mass eg) wih By vol Yas _—_—ca ow a) NaLz S00 13 1560 ear Bt 1s) aaa ‘Ae noted svete bef ofthe locks was immersed in ana of water the ‘pected eutaton temaved an then wiped wits dap dth pro bonding “he bende boc were taped dwn to cele ah S12 mm dick ein Joie aces were mage Bish wt he Backs ec] The euplets were reaced {eC indr meted lack werebended win ta er imerson tot Mh cerng ene replicated oth shekeed site onions nd commen cng bracce bonded sandstone couplets and oa pt speiners were laced bt ‘eloventiated setecedottdoorenronment and complete ered under Dolvnene teraz to ensure sfe Hari er al fdr se Aer = {iy the polythene sheeting and tata sample tous were resaved for the eimaing 21-dys, Daly mi-aycrmperatresCaned between 15 ad 20°C 25; Mortar sent and deformation eremiaed by tee point long on 6,40 = 40 T60%nmh pis specene In Secordance wh BS EN 1015-11 1999 [22 using & 100 faston A208 testing Imacine The comprenive sueng of the mata way then eetermined on the {ho pars resulting om the Hexa stegth Lest. To study the mechani be ‘out of the mora under lan, Nena ane compressive loe-erormation profiles ‘were recorded to fce and subreqently converte to ressstan caver, 26 Many on renee “The bond-srench meted of establishing masory exural bond stent 2) war acopte inthe sy rater than te wal pane! Renal tert 2>| hasbeen ‘how bowever tht bet tess produce simlarresls[5 24 inthe bonerene inetd, a musty stackSonded piss ofa coupe is subjected fare which wrenches the uper lok apart om the ite ower {ices the mora bed join “The appara is shown schemacially ny 2 ane essen allows that Aesebed in HS ENI05I-5.2005. The eckmarr eupiet was longed to fre Sy ineementaly isreasing the mass ofthe lever sm 8200 internets Te ode of tlre wat recorded andthe bond strength of masonry atl fo the ig. 1. Cull sandstone couplets approxitely 5 min afer Bonding Tas ute lf shows the postion ef the waterston ithe sandstone uns afr those us ‘within resetting afr hase its with Ora pre-weting The water- lines. In(b) the wate-rent wl represent the water bsored om the marta bee ig. 2. Schematic diagram of renter Devious ofthe mortat jo. With reference te the Bond Sef aalute ieiwenby, fo BETAS “HUE +28 hee Z=bi6 Bs themes wit of he bed ine che mean eet the specimen an the stance om the appli loa the esin face othe Specimen me, he dsance fom te cates gravy of the camping systems From the censin ace f the specimen (nF se app lod (the Weight of the bond wrench spparats (Nd Wis the weight of he saan at Palle he specien (N),together with any adberest mortar. The characteristic Serene as sbsequent evaluated based on «95% conidece lvl of og 3. Results and discussion 2.4, Mortar strength and deformation ‘Mean mortar strength, standard deviation (SD) and coefficient of vatiation (CoV) are presented in Table 2 for the four binder types. Iti evident that binder hydraulicity had a profound effect on both flexural and compressive strength, with strength increas- ing with increasing hydraulicity. The presence of lite, calcium alu- rminates and elevated belite phases within the NC mortar has resulted in rapid strength gain within the 28-day cure period. ‘The high belite content of the moderately (NHL 3.5) and eminently (NHL5) hydraulic limes has also resulted in strength gains over the feebly hydraulic lime (NHI 2). The presence of free lime within the NHL mortars, particularly NHL 2, leads to slow strength gains as hardening is primarily due to carbonation and the extent of sable ‘Story f strength test reutsfr marta binders, Tinder Renal angi ‘onapesive ent 97° Man sD CG Mean SD (hers) fea) Gera) WH? 0198 00199430358 om9 8a NHI35 0255 002 $75 G90 G70 aA Ne S507 089 Ges aes ago Ftcarbonation over the 28-day period would be limited, The com- pressive strengths are in the range 20-25 times the Mexural strength. The mortar strengths obtained are generally comparable to a study (5) utilising similar mix proportions and curing regime. Tis apparent that the NHI. binder designation overstates the ‘mortar compressive strength achieved in practice, This is due to the 1:3 mortar mix ratio (binder:sand by mass) prescribed in the building lime mortar classification [25] and strength testing codes [26]. Specifying a mix ratio by mass for relatively [ow density limes results in an overly rich mix generally not used for mortars in prac- tice. Also, itis important fo note that the strength of the mortar Within the masonry joints i ikely tobe different from that of mor- tar prisms due to factors such as masonry suction reducing the water content ofthe mortar and the reduced exposed surface atea affecting carbonation Flexural and compressive stress-strain results are presented in Figs. 3 and 4 for the four binders, Each Figure presents the com- plete curves for five, notionally identical mortar samples. From Fig, 3(a)-(4), it i evident that, in flexure, all mixes strain linearly until fracture, which occurs suddenly with no plastic deformation. Although the flexural tensile stress at failure increases mono~ tonically with increasing hydraulicity of binder the strain at failure generally lies in the region 0.003-0.004 mmjmm. Based on the slope ofthe linear portion of the plots the sifiness range obtained for each binder type is presented on these figures with stiffness increasing with increasing hydraulicty ‘With reference to Fig. 4, under compression, the binders display an initial linear-elatic region with deviation from linearity gener- ally ying in the range 0.0075-0.01 mm/mm strain. Unlike the flex- tural response, plastic deformation is evident; however. the extent of plastic deformation decreases with increasing hydraulicity of the binder with the NC binder displaying a well-defined descend- ing branch to the compressive stress-strain curve over the stress range presented. The stiffness range for the binders obtained from the plots is displayed on the respective figures and are in general agreement with those obtained from the flexural test, 45.2 Cullalo sandstone water absorption rate ‘The capillary absorption test results for the Cullalo sandstone are presented in Fig, 5, with data plotted on a square-raot-time Axis: Fir S(a) presents the test rests taken aver a 72-h absorption period and Fis, 5(b) presents the curve over the inital 1-h absorp- tion, The error bars on the data markers represent one standard eviation on either side of the mean and where the error bar appears to be missing, the marker is larger than the error bar. The slope ofthe inital linear portion of the graph gives the coeffi- cient of water absogption due to capillay action which was obtained as 305 gms”). Comparison of the tate of absorption of Cullalo sandstone with other types of stone was not possible as published values are scant and is due to the rate of absorption being a non-mandatory test. Itis also difficult to compare absorp- tion rates of Cullalo stone with blocks of other materials to BS EN 772-11:2011 as there are different immersion durations and units fof measurement for different block materials. Supplementary absorption rate tests showed that one sample ofa clay engineering brick and one sample of a high alumina concrete fire brick have. respectively, a coefficient of water absorption due to capillary action of 20 g(m?s°*) and 900 g/(m’s?®} wien measured in the same manner as natural stone (Fig. 9(c)}. This indicates that éry Cullato sandstone has a moderate rate of absorption. ‘The immersion times of 1 min and 15 min are bighlighted on fig, S(b) and show that at these times, St g and 125 g of water (respectively) had been absorbed into an individual sandstone block prior to bonding. This represents, respectively, 1.6% and 4.0% absorption ofthe dry weight ofthe block or 33% and 81% ofthe total ‘water absorption at 72 h. The overall porosity ofthe sandstone spe- cimens was evaluated as 11%. With reference to fig. 1 this figure also highlights the position of the water-frontin the sandstone units Which have been subjected to O-min and I-min pre-wetting. The water-front for O-min will represent water absorbed from the mor- ‘ar bed-joint during the period after bonding and when the image ‘was taken, which was approximately 5-min. Low E: 47.56 MPa Basmo Mee I | "I JE = ig 3 ler stesestrin crv fr five mortar prisms (8) NHL2,(9) NHL, () NLS, ae 4) NC Note: ities range denoted Eon Figureha 5 Bu ea /Comsruction and Bulding Material 842015) 128-135 : ° oun ss a E22sMPa E62.9sMPa +E: 75-95 MPa (one E:ne620MPs Fig 4. Compressive sss-stan curves for ve mrtar prisms) NHL 2.(b) NHL 35, (€) NAL. and() NC. Noe: sis ange dented Fon utes - w , 2.3, Flexural bond strength The mean (fu) and characteristic (fae) bond strengths of the masonry couplets are presented in Table 4 for the three bed-face pre-wetting times viz. 0, 1 min (1.6% absorption) and 15 min (4% absorption); the standard deviation and CoV of the test results Bre also presented, Considering the characteristic strength results, all NHL masonry tested at 28-days ranged between 005 andStar ea /Comction and Buin Materia 84 (2015) 128-135 ‘able Summary of Band strenght fr dierent prestige ye imean) DG gion) piers) SDV J (mean)Jolmesn) GeV Jue) Ms) (ea) ie) (Gera ra) (3) Gara itera) (shea waa o107 ows 79 oes 006 oan 38 om” ——o6e von 1840050 NHLSS 0233 00g 2a oss ms 27 OS tas bo "sa o156 id nt fal aan ond on bend-wreach appa 1 2 1 1 la? Fg 6. Fare oferta bee joint showing a) esion failure at iteace between o mera 8 upper block, and (b) agonal tension fire within mortar besoin Nate {= tension face 2s compression fre scemac grams tna) an 0 aged fam 211). (0.31 MPa indicating that they all complied with the value quoted for natural stone masonry in the masonty design code [3] Le. bond strengths 0.05 MPa for failure parallel to bed-joints. Characteristic strengths for the NHL 2.5 and NHL 5 masonry at 26-days (4% absorption) were 0.16 MPa and 023 MPa respectively, which is greater than the value of 0.15 MPa given in the specification of, ‘masonry mortar code [27| and comparable to the 0.2 MPa strength quoted for M2 cement mortar bonded to concrete and calcium sili- cate blocks in the UK annex to the masonry design code |. Mean bond strengtins of all NHL masonry tested were within the range 0.07-0.33 MPa, This is lower than the range 0.2-0,61 MPa reported by Pavia and Hanley (6) utilising a richer 1:2.5 mix ratio (by mass) but within that of 0.05~0.63 MPa reported by Zhou etal. 5) utilis- ing a longer curing duration (91 days); i also lies within the 0.11- 0.16 MPa range reported by Costigan and Pavia [7] utilising the same curing duration, An additional possible reason forthe greater strength results found in previous studies (5.6 i their use of per- forated bricks allowing mortar to flow in and bond to the perforation sides which may have led to overestimated bond strengths due toa greater bond area than that assumed, The failure strength of all NC samples was undetermined as a bond stress of, 1.09 MPa - the limit of the testing equipment ~ did not induce fail- ture. The upper limit of expected bond strength is assumed to be approximately 1.5 MPa as found in mortar flexural strength test- ing. However, the value of 1.09 MPa bond stress isin excess of even the highest characteristic value given in any masonry code of prac- tice; the national annex to BS EN 1996-1-1:2005 gives a value of, 10.7 MPa for low absorption clay masonry bonded by a mortar des- Jgnation (i) of strength class M12 (the mumber following the letter Mis the compressive strength forthe class at 28-days in MPa), For the NHL mortared couplets. two failure modes were observed: the NIL 5 samples generally failed at the interface between the mortar and upper block shown in Tig. (a) whereas the NHL2 and NHL 35 couplets generally failed by tension failure diagonally across the mortar bed joint shown in Fis. 6(b). No cou- plets failed due to failure of the block As noted above, the NCbe 5 bey ea /Comruction and Bling Material 842015) 128-135 couplets did not fail under the maximum load which could be applied to the wrench apparatus The results ofthese tests indicate values that could be achieved in practice asthe specimens were subjected to realistic site curing conditions over a 28-day period and using realistic mix propor tions, albeit sheltered from rain, Previous studies on NHL mortars [7] have found substantial increases in bond strength between 2B-days and 6-months therefore the values presented may thus represent less than half the expected long-term bond strength. ‘The work presented in this paper has shown, however, that itis still practical to use 28-day data despite NHL mortars slow strength gain, 24, Block absorption and bond strength Fig. 7 presents the relationship between bond strength and block pre-weiting time, in terms of percentage absorption of dry ‘weight It is evident that increasing pre-wetting time of the dry Cullalo sandstone block results in decreasing bond strength: fur- thermore, as the pre-wetting time increases, the scatter in the results, aS quantified in the CoV (Table 4), inereases for all NHL mortar types. It is anticipated that this is related to the inherent variations in porosity through the individual blocks, particularly near to the bed-joint surface of the sandstone block. The evidence for this can be seen from Fix. 5(b) whieh shows thatthe standard deviation of the cumulative absorption at 15 min is greater than ‘that at { min In addition, the influence of pre-wetting on the bond strength becomes less significant with increasing hydraulicity of ‘the binder: for example, considering the mean bond strength with ‘min pre-wetting. relative to the dry block, bond strength this is reduced by 29%, 17% and 5% for, respectively, NHL 2, NHL 3.5 and NUL. Guidance regarding pre-wetting of masonry units is limited in current execution codes of practice. BS800-3:2001 [28], for exam- ple, advises that stone units should have tral courses built dry to test bond prior to possible wetting whereas BS EN 1996-2:2005 [25) simply states that the specification should be consulted and, if there ts none, to consult the manufacturers of the blocks and mortar. The results obtained inthis study indicate that blocks with moderate absorption rates of <:300 gi(m* x 54) should be kept as dry as possible for optimum bond 135; Mortar strength and masonry bond strength Fig, 8 presents the relationship between the compressivejfex- ural strength of the mortar and masonry bond strength. Unlike mortar flexural strength, compressive strength exhibits an almost linear relationship with bond strength. This would imply that. for oo orp ie 7 nuence of pesmeting on sands Fe 4. tauene of mortar strength (Flexural, C~ compressive) sandstone band scent fr 0% 18% and absarpton (peowecing the NHL mortars, compressive strength would be a good indicator of Cullalo sandstone bond strength, According to the UK National Annex to Eurocode 6 |4), cement mortars with 2 compressive strength of 12 MPa bonded to low absorbency clay bricks have a masonry flexural strength of 0.7 MPa with the plane of failure parallel to the bed. joint. Considering that NC mortars have a similar total proportion of hydraulic components to cement mortats, their compressive slvength is lower than cement mortars yet their bond strength is higher. The bond strength of NC mortar may therefore have benefited from the high calcium aluminate content which brings a rapid set and strength gain resulting in optimal water transfer for the absorption rates experienced, 4, Conclusions and concluding comments This study has presented both mortar strength and sandstone bond-strength data using practical mortar compositions, stored and cured under natural conditions - an area where there is, cur= rently, a dearth of information, The work would find application in developing guidance and specifications when such materials are used in conservation, restoration or refurbishment work and also where structural assessment is required. The following can be diawe from the investigation! (1) tn flexure, all the mortar mixes strained linearly until failure Whereas under compression a plastic region was detectable, the extent of which decreased with increasing hydraulicity ‘of binder, Mortar stilfness increased with increasing hydra licity of binder. (2) The bond strengths of NHL 3.5 and NHL § mortared stone masonry after a 28-day cure were observed to be compara~ ble to low-strength cement martared brickwork, The bond strength of NC mortared stone masonry at 28-days exceeded ‘even high strength cement mortared brickwork. (2) Block pre-wetting had a significant influence on flexural bond strength of sandstone masonry, with bond strength decreasing with increasing pre-wetting time. The results of this study indicate that for optimum bond, blocks with mod- ‘erate absozption rates of ~300 gi(m*s®*) and under, such as the Cullalo sandstone tested, should be kept as dry as possible. (4) Mortar compressive strength had 2 profound effect on the Aexural bond strength of sandstone masonry exhibiting a positive linear relationship; as a result, flexural bond strengths of stone masonry should continue to be cate- _gorised by mortar compressive strength in design codes of PracticeStar ea /Comction and Balding Material 842015) 128-135 Bs ‘The work would also indicated that mortar strength testing. and NHL binder designation codes of practice 25,26), should relate to imix ratios commonly used in practice (as this study has used). for example, a 1:25 mix by volume rather than the 1:3 mix by mass. Despite the slow strength gain of NHL mortar. the work has shown that it is stil practical to use 28-day data as prescribed in current codes of practice [26] but the strength obtained may represent less than half the expected long-term strength [7]. The wark presented will also serve to promote awareness of hydraulic lime mortared sandstone masonry and to enable more confident design and assessment of this material Acknowledgements ‘The authors wish to thank Falkirk Council for accommodating the study: staff a the Scottish Lime Centre Trust for their expertise and use of laboratory facilities; Masons Mortar Ltd. (Edinburgh) for supply of materials and Mr. C. Edington for his assistance with the bond-wrench apparatus References 1n)sroctebank ear. sung lies in conservation. Donhead Publishing 2012. p36, SN 9781875591953, [a] fart Aan assessment of the eationshp Detween the water vapour Pettey and hyraly of ie based marcas with arc reltence fo bulleing conservation materials scence PRO. Meno Wat UnVeSI oon 1b) rsh standards nsaton (8) Bs EN 1986--1:2005%A1-2012Furocede "esgn at masonry srotares ~ ar tt general Ts for reinforced ae Sorenfored marony stacre: BS london [a] nts stances nstaton 8S Naonal Annex to BS EN 1996-1-1:2005 UK National Aanex co Farorde 6 design of mason structs = Pat T-1 [ene efor rehofree and unenfreed masony trates BS Loney Is) how 2. Willer P Daya D. sength characte of fda ime ‘ortare bkovork: Proc ino CEng Cant Mater 2008°1516)159-46 [oy Paves Haney Fear! bond sength of sta pera ime maar sed 10] Cogan A Pavia lnueee ofthe mecha] properties affine err on Se sng fb sey I Wk Sage. at er {epue New Yk Springer 2013p. 359-72 Is) Lawneace M Wallet P2howZ Inuenc of aetacial material pore structure fn the senge ofthe bine morta bord dem, p373-381 [a] Henin Reels S len KY. Water tanspert between sna rick: the Jngece of material arate Ide, p. 329-341 110) tana. alarer Masonry rep ime-ased mortars factors aectng the ‘chanical bent Cent once Res 20032311) 1867-75 lot) Mtghes Vale Moras in hse Daeg: review the scenic and fonsereton Iteitue store Sctland: 200.97 (SBN 1903570823). tna] Venknrana Reaey BV. Cupia.Aindoence of sand. gracing onthe Characteristics of tortats and sol-cement lock masonry Cons Bald Mater 2000218): 161425. 103) Kage A Karta | ig Diner on micrroctre and Sud Mater 20145312) a6)-78 [4] Meciniey WM IRA an the earl bond tenth fly brick maton n Matis edie. Masonry coropenecs to asserbiages. ASTM Spec [iS] Rtehe T Davison Jl Facorsafeting bond strength and resstance to [ASM Spec Techical Peton Na 320186, p 16-20 106) Se: rwwrtadstosctsiercesiconnypredution?> — er “indtone blac production deta [acessed 250213 [u7) Bat standveststtin (BS), BS EN 772-11 2013. Methds af tse for ‘no PT arian of nate apn of sete ‘masony ats duetoeapllry ation and the alate a ate absorp at lay maar ut St Londen (19) Stanamineral Usa foe 2014 Sut Are atu hydra ine ~ mero {chemistry of raw mater dvalable a hp on linesuatece ‘Romine aloy-nensteyataateca cesed 2502.15) [19) SemmalnD. Teena specications ~ promt natural cement = the Roman ‘cement of Grenoble. Avaliable at «por emanporiand nets ou) {Geechee engp 209 (acesed 25 0213 (20) sees standares sition A) BS EN 101531099, Metbeds of test for ‘mora fo masomy = Pat deterninaon of consistence of esh matt (> fw table! BS London [2t) Bish Snead tsoton 85 EN 1082-52005. Methods af test for masoney Trae 5 deteraton of bond strength by the band wrench eto BS. Tondon [2a] fre Standars sion, SEN 1015-11-19, Methods fet for mort formacony= Par eetermination of fea nd compressive eng of fardened marar BS London [23] Het Standards etteton 85 BN 1052-27899. Methods a tert for mtanny “pared termination of fives Stengh lender [2 lig Resch sient (BRE HE ge 260 — Ts [2s] Bitak tandaasotiton 6S EN 459412010, Bling tine Pare 1 (26 Bio Sts tun SN 4S. 2010. Bln ine Pt 2 tet [a7] tah sundae nsctuton, BS EN 998-2:2010. Speciation fr era or Imusony = Pst 2. many mera BS. Landen [ae] utah edad ntti, 26800022001, Workmanship on balding sites put: code of practice fot tases 5), Lode [ao] dota Stancude Instn, EN 19962.2008,Euroode 6 = Design of ‘masohy stitutes ~ Put 2: desig cobsideraos, seecton of atl id V. Te elect of arena size and type of chanical properties of HAL eras. Cons bone