You are on page 1of 17
The Smeaton Project: Factors Affecting the Properties of Lime-Based Mortars JEANNE MARIE TEUTONICO, IAIN McCAIG, COLIN BURNS, JOHN ASHURST ‘The broad objective of the Preface ‘There is much discussion regarding Smeaton Project isto contribute 2 The frst phase of the Smeaton this test and freeze-thaw durability the understanding of the character: Project was intended ro bea eange- fa” enero! on 2 Suropean and isties and behavior of ime-based finding exercise, which would decided to utilize the sae crystalliza- ‘mortars for the repair and conser- 1 sis the project team in refining ign gest Largely to asses it stabil Vation of istorebulldings The emethodswsed for preparation, yd io etablsh some bass or article presents the first phase 2) establish some trends so asco ber. CO™mParvon of samples. Modifica. tions have been made in the experi- results of a joint research program ter define and limit future phases of Thenral design of Phases Il and il ofICCROM, English Heritage, and ‘he research In general these Objec~ jase on this experienc Bournemouth Universi ives were achieved. Finally it was impossible, in his a However, the authors wish to very preliminary stage of research, to ‘emphasize the very preliminary consider all variables that might have nature of the findings reported in this HN she behaviee of the ceortas paper. They are published primarily tested of to explain all mechanisms to promote dialogue and, pethaps, observed, This is the nature of scien- more widespread collaboration in the fife enquiry. It ig hoped that some future. Inno way should the results uections may be anewered through be considered definitive or con- tore detailed laboratory studies later lusive, in the project as well as through The paper should, thus, be read as Gomparison with long-term results an exploratory exercise that will be observed on the exposure site and in used to focus and refine future work, Feld vesting. rather than as a conclusive scientific study. In the first phase, insufficient samples were produced to statistical- ly validate trends. The consistency of results was also affected by both the method of preparation and cure of the samples. Problems with demoulding and cutting of the large samples used in Phase I, for example, meant that certain samples were not available for all test procedures. Based on this experience, both prepa- ration and cure methods have been altered in Phase Il, which is now underway. Similarly, there is controversy about some of the test procedures utilized. In particulag, the salt crys- tallization test that is stil widely used to assess durability inthe U.K. is no longer recognized by the Canadian Standards Association and is currently under review by ASTM. Introduction Interest (or perhaps renewed interest) in the use of lime based materials — mortars, grouts, plasters, renders and paints — for use in the repair and maintenance of historic buildings and monuments has been growing steadily in the international conser- vation community for the past 15 to 20 years. ‘To a large extent, current practices have evolved through trial and error informed by only limited scientific and academic research. Although there isa significant body of experience in the use of lime-based materials in certain parts of the world, itis apparent that practice is not always matching theory and that there are still many partially un- 2 solved problems. ‘The characteristics of mortars may be defined in several ways. In practi- cal terms, the ones that concern us most are: 1. Of the fresh mortar workability moisture content rate of hardening shrinkage 2. Of the hardened mortar + appearance ‘moisture and air content permeability mechanical properties including + adhesion + ability to tolerate ‘movement + strength + durability (resistance to dam- age by frost and salts) ‘These, together with the chemical properties of the mortar, must clearly be compatible with existing historic ‘materials and appropriate to the con- text in which they are to be used. ‘The factors affecting the charac- teristics and behavior of lime mortars are derived not solely from the mor- tar constituents but also from the techniques used in processing the ingredients and in preparing and placing the mortars. The properties of materials in contact with the mor- tar and ambient environmental con- ditions at the time of placing and hardening of the mortar are also influential. In summary, there are probably three categories of problems that need to be addressed. These are : 1. Mortar analysis and the com- ponents of historic mortars, 2. Performance criteria and com- ponents of specification mortars 3. Mortar preparation and utiliza tion ‘The Smeaton Project — a joint research program of ICCROM (the International Centre for the Study of the Preservation and the Restoration of Cultural Property), English Heritage (the Historic Buildings and Monuments ‘Commission for England), and Bournemouth University — was set up in response to the need to address such issues in @ systematic way through both laboratory and field research, ‘The broad objective of the Smeaton Project is to contribute to the understanding of the characteris tics and behavior of lime-based mor- tars by attempting to identify — and where possible quantify — the mate- rial and practice parameters that affect their properties. Specific emphasis has been placed in the first stage of the research on material and practice parameters that will improve the durability and frost-resistance of lime-based mor- tars in harsh climates. Itis hoped that the research will ultimately lead to the production of practical guide- lines for specifying, preparing, and utilizing conservation mortars in a wide variety of regional conditions. The name of the project refers to John Smeaton, a mathematical instrument maker and engineer, who in 1756, after experimenting with limes and additives for “water build- ing” (mortar capable of hardening under water), decided to lay the courses of stone for the Eddystone lighthouse (20 miles off the coast of Devon, England) in a mixture of pozzolana from Italy and Lias lime from the southwest of England. His report on this work! was the first of many studies preceding the current project that, ina sense, brought together England and Italy in seek- ing to identify and quantify poz- zolanic additives for optimum per- formance. THE SMEATON PROJECT 98 BACKGROUND TO CURRENT PROJECT ‘The Smeaton Project grew out of experimental work begun by English Heritage in 1986 to identify suitable mortars for use in the conservation of Hadrian’s Wall (Fig. 1) in the north of England, Sections of this Roman wall are on high ground and ‘exposed to severe weather condi- tions. In the recent past, relatively strong mortars gauged with portland ‘cement had been used for pointing and repair. This mortar was itself sufficiently durable to withstand the ‘extreme exposure to which the Roman wall is subjected but con- tributed to the deterioration of the wall in a number of ways that are typical where cement is used in con- tact with weaker and more porous materials. ‘These deleterious effects include: 1. Concentrating cycles of wet- ting and drying through the stone faces because water cannot escape through the relatively impervious pointing mortar 2. Trapping water, thereby caus- ing leaching of original core morcar and increasing the risk of frost dam- age 3. Increasing the likelihood of mechanical damage to stone on removal of high-strength mortar in the course of maintenance works Recognition of these problems led to an attempt to exclude all cement in mortars used on the Roman wall. However, early trials with lime mor- tars were not completely successful because the mortars had inadequate frost resistance. While their behavior cally satisfactory, such mortars had to be replaced at a frequency that ‘was not economically sustainable for English Heritage. 34 Aer euLLETIN Fig. 1. Hachian’'s Wal, general view. Mortar Triale at Hadrian's Wall A testing program wat thas initiated to develop more appropriate conser- ‘vation mortars forthe site. As a first step, a number of oint- ing and core moctars were sampled from areas of the Roman wall known to contain original material. ‘These were found to contain lime, crushed tile, crushed sandstone, ssand, and kiln debris. The binder: aggregate ratio ranged from 1:1 t0 13. Traces of animal fat, probably tallow, were also observed in some of the samples. ‘The next phase was a program of comparative durability tests on a range of 120 mortar mixes com- ‘menced at Hadrian's Wall in 1986, ‘This exercise was designed primarily to observe the comparative perfor- mance of the binding materials in the mortars and was not intended at that stage to simulate site practice. The trial mortars were grouped in three categories according to the type ‘ofthe principal binder material: Group A: non-hydraulic limes (both putty and dry hydrate) Group B: hydraulic limes Group C: portland cements ‘The mortars were mixed in vary- ing proportions of binder: aggregate with and without pozzolanic and air- centraining additives. Principal por- zolanic additives were ordinary brick: dust (essentially crushed facing brick ‘obtained from s local manufacturer) and HTI Powder (High-Temperature Insulant-refractory brick dust), both. crushed to passa 300 micron sieve. ‘The sand type and grading was con- sistent throughout the tral mixes and conformed to relevant British standards for building sands. ‘Two 150 mm and ewo $0 mm cubes were prepared of each sample in the spring of 1986. After an initial curing period of two weeks, the larg- er mortar cubes were placed on exposure racks (Fig. 2) in two loca tions near the Roman wall. The two smaller sets of cubes were stored in a ventilsted frost-free environment and were subjected to the British Research Establishment (BRE) sal- crystallisation test? in spring, 1987. Fig.2. Mortar tial, Hedran’s Wall ‘The samples exposed on site were visually assessed for signs of break- down after each winter period, utiliz~ ing simple A - E code to describe the degree of deterioration. The leter A described mortars which showed no deterioration; E indicated rotal disintegration. ‘The letters B, C, D and various letter combinations were used to designate levels of deteriore- tion in between these two extremes. In 1988 selected mortars were used in the consolidation of small trial sections of the wall. The purpose of these trials can be summarized as follows: 1. To evaluate the practical issues involved in the preparation of lime ‘mortars both on site and at a remote depot 2. To evaluate such issues as workability and ease of handling in the use of lime mortars on site and to record any problems encountered 3. To provide realistic and varied test sites forthe long-term moni- toring of performance of the tial ‘mortars ‘A joint inspection of the exposed ‘cubes end of the tral sections of wall in 1989 posed many questions but also indicated trends, some of them surprising given assumptions about lime mortars and some stan- dard conservation practice atthe time, Similar trends were observed ina report produced afte a five-year ‘exposure period,* which compared field results to those observed in the laboratory tests. These trends include: 1. In the mortars based on non- Inydraulic lime, the best performers were the mixtures of lime sand:brick dust in proportions 1:3:1 (Fig. 3, samples A2B, A2D). In all cases, the standacd brick dust performed significantly better and more consistently than the HT Powder (refractory brick dust) 2. The worst performer wes the mixture of lime and sand, to which a 1/10 part of white cement had been ‘added (Fig. 3, Sample A3D). In fact, all ofthe lime:cement mixtures performed poorly until at least 1/2 pact of cement had been THE SMEATON PROJECT 38 added to the standard 1:3 mix This trend was a bit disturbing, given the widespread tendency in the UK to add small amounts of cement to lime mortars in conservation ‘work. However, the observation did seem to be consistent with laboratory results that had been achieved in an earlier Swedish study published by Ingmar Holmstrom in 197.6 3. In the mortars based on Iydraulic lime, results were extrome- lyinconsistent, depending very much on the type of hydraulic lime utilized inthe mix (Fig. 4). ‘Those made witha buff hydraulic lime (Samples BI A-D) performed reasonably well; those mixed with a white hydraulic lime (Samples B1 E- H| showed severe deterioration. In the case of the buff hydraulic lime, performance was markedly improved by the addition of one part standard brick dust (Samples B3 A-D). And again, the standard brick dust was ‘more successful in improving the durability of these mortars than equivalent parts of the refractory brick dust (Samples B2 A-D).. Clearly, there was a need to try to ‘avantify some of these empirical observations and to determine the cffect of various factors on perfor- mance, The overall framework of the Smeaton Project was, in fact, designed in response to the questions raised by that site inspection and the frustration inherent in trying (© specify materials based on the observed results. ‘THE SMEATON PROJECT ‘The Smeaton Project isto be carried out in phases that will test a variety of material and practical parameters over a five-year period. A general outline and schedule for the project is given in Appendix A ‘Scope of Work Given the questions raised at ‘Hadrian's Wall and in the subsequent literature search, it was decided to focus the initial phases of the rescarch on the effects of set addi- tives, specifically brick dust and ‘cements, on the performance of limessand mortars, The first phase, completed in ‘August, 1992, had two main objec- tives 1. To confirm trends observed in the field and in the lierature regard- ing brick dusts and cements 0 as to develop parameters and limit vari- ‘ables in future phases In terms of brick dusts, an attempt ‘was made to understand the effects of such factors as optimum particle size, firing temperature, and propor- tion of dustin the mix. In terms of cement, it was hoped to confirm the trends shown at Hadrian's Wall and in the Swedish experiments indicating that small additions of cement to lime:sand mortars actually redvce the pecfor- ‘mance of the mortar. 2. To develop appropriate meth- ods forthe preparation and testing of samples to obtain useful results regarding the established variables. Since most mortar text procedures ‘have been developed for cement- based mortars, this was an important area of investigation Phase! ‘Materials and preparation of samples. A series of 34 mortars were tested in Phase I (Appendix B). All were based on a coarse stuff of one part lime purty to three parts sind by vol- tume, modified by the addition of brick dusts and/or cements.” In gen- eral, the mix types fell into three categories: Group A: based on lime:sand coarse stuff plus various amounts of four different brick dusts in particle sizes from 75 t0 300 microns. Group B: based on lime:sand coarse stuff plus varying amounts of Fig 5. Somple moar block, Fhe either white ordinary portland cement of sulphate resisting cement. Group C: based on lime:sand coarse stuff plus brick dust and white ordinary portland cement, All tese mixes were prepared by English Heritage. Initially, lime putty, obtained by slaking lump Time, and sand were blended with ‘excess water in a “Rollpanit” roller pan mixes. The resultant coarse stuff was then stored for a period of to months in humid conditions. At the time of sample prepara tion, the seasoned coarse stuff was reworked using a small "Mixal” ‘mixer and by hand. Brick dusts and ‘cements were added in the form of a slurry at this stage. ‘Where necessary further mea- sured amounts of water were then added 0 the mixes before moulding to achieve a consistency suitable for bedding as judged by English Heritage master crafismen. “The prepared mixes were then, hand compacted inte wooden moulds to create large rectangular mortar blocks (600 x 100 x 100 mm) (Fig, 5). These blocks were removed from their moulds after one ‘week and then placed in an environ- mentally controlled room at 25° C and 90% RH for 120 days. THE SMEATON PROWECT 97 After this curing period, the blocks were cut to size for the var: ‘ous tests which were carried out by the Weathering Sciences Section of the Building Research Establishment under contract to English Heritage. Test procedures. The following tests were performed: 1. On the fresh mortar: ‘Moisture content. Samples of cach wet mix were placed in ¢ small plastic bag and weighed. The sample bogs were then slit open and placed ina drying oven at 60°C until con- stant mass was achieved. The differ- ence in weight was vsed to determine the moisture content of the mix, expressed as a percentage of the mass of the oven-dry sample. Stiffening rate. This was deter- mined at intervals of 15 minutes, ewo hours, five hours and 24 hours after placing using a Stanhope Seta pen- etcometer (Fig. 6), The penetrometer ‘measures in 0.1 mm units how fara ‘weighted cone penetrates the mortar 2. On the hardened mortar: Compressive strength. Two 100 ‘mm cubes were cut from each prism and tested on a standard 200 kN ‘compressive test machine, Before testing, the samples were immersed 38 APT BULLETIN Fig. 6. Ponewometer used 10 st sttfenig rate of fresh mortar in water at 20°C for at least 24 hours. The load rate was 0.03, Nimm?s. Water vapor permeability. This was measured using a modified ver- sion of BS 3177:1959. Samples measuring 25 x 100 x 100 mm were placed in aluminium trays containing, approximately 70 grammes of calci- tum chloride. ‘The edges of the sam- ples were sealed to the sides of the ‘ray with hot wax, leaving only the 100 mm? face free to transfer water vapor. The samples in their trays ‘were then weighed and placed in a relative humidity cabinet maintained at 75% RH. The increase in weight ‘was measured and plotted against time until no further increase in weight was observed. Depth of carbonation. This was established by spraying the freshly ‘cut surface of the sample with phe- nolphthalein (a standard pH indica- to:). The sample turns purple when the alkalinity ofthe sample remains ‘greater than pH9 to 9.5 leaving the ‘carbonated areas colorless. Sodium sulphate erystallization test. This isthe standard test used by the BRE to approximate freeze-thaw cycles and provide an indication of the durability of a stone or mortar. Four 50 mm mortar cubes of cach sample were subjected to 15 cycles 1, soaking in a 21.5% sodium sulphate solution for two hours 2. drying in an oven for 16 to 20 hhoars at 105° C. ‘The weight loss after the 15 cycles ‘was used to rank the different mor- tars. Fig. 7shows the results of 15 Fig. 7. Semple cubes ater 16 cycles ofthe salt srystlization tet. cycles of the sale crystallization test ‘on one of the samples. In addition to these laboratory tests, 33 cubes were placed on a rack on the BRE exposure site. Before exposure, samples were dried in an oven at 60° C to constant weight and the physical dimensions measured. It is intended to measure and weigh the samples at intervals of six months (Fig. 8). Discussion of resuts ‘Moisture content: Mortars had a relatively consistent water content, in range from 15 to 19% (Table 1). Above all, this result is an indication that the sli ofthe craftsman is an important facto: in any conservation operation. There are no obvious trends, other than a general reduction in the amount of water added to the lime:sand:brick dust mixes as the brick dust size is reduced. This may be due to the fact that the larger brick dust particles tend to function as porous particulates (which absorb Table 1. Moisture Content Sorted by increasing content Sample Moisture Sample Moisture No. Contes No. Content ag 181 al 17.02 al 1525 aS 1707 byt 18340 an 17 byl 15620 aM 7M ad 1874 eat ATA ai 1621, a3 1727 ait 1622 a7 38 al 164771740 ait 1655 ba 17.42 aM 1661 adsl 17.54 ail 167223761 asl 1674 ba 17.82 cul 1674 ak 17.80 bat 1681 aa 18.27 a24 1681 aL 18D Ql 1689 a9 18.84 an 1691 A201 13.30 water in mixing), a factor which will be discussed further in this paper. Stiffoning rate. Weis dificul eo discern subtle trends regarding stft- ‘ening rate, due to the limited number ‘of samples tested in this first phase 260 240 220 200 180 160 140 120 100 80 60 40 20 o Fig 9. Stlering rte cuve: Samples A712, and the wide number of variables. However, certain general trends can be observed Regarding the lime:sand:brick ddust mortars, there were wide varia- tions in the depth of penetration of 20.00 THESMEATON PROJECT 39 the mortars after 15 minutes, proba- bly due to the variation in water con- tent and working. ‘The mortars made with one of the lower fired brick dusts did tend to harden slightly more rapidly than the other lime:sanc:brick dust mortars. Inall of these mortars, however, the shape of the curve is relatively similar and probably representative of the numerous factors which leed to stifiening in lime-based mortars (Fig. 9). Initially, the mortar simply tends 10 revert 10 the stiffness it pos- sessed before re-working on site. Subsequent stiffening is caused by loss of water and, erentually, by carbonation. By comparison, all mortars from ‘groups B and C (containing cement or cement and brick dust) hardened aker two hours. Iris interesting to note, however, that this initial speed in stiffening bore no relationship 0 the eventual strength and durability achieved after 120 days of cure. Carbonation depth, ‘This test was not expected to establish trends, but simply to provide some idea of the values which could be expected under Table 2. Carbonation Depth Sorted on increasing depth Sample Carbonation Sample Catbonation No. Depsh (mm) No. Depth (men) byt 0-1 ats 100 byl 80 an 110 al Oa 110 bat 60 at 0 an men 110 ae a7 11.0 bu ba ao al0/1 cil 120 aan asf 120 alll ale 13.0 av BO 10-17 an m2 140. 4/1 as 130 ay a1 15.0 aul by 130 abt a 170 aia 40 APTBULLETIN the established curing conditions. As seen in Table 2, carbonation depth varied from 0 to 17.00 mm. Compressive strength. Here, the tests (Table 3) confirmed the trends secn at Hadrian's Wall and in the ‘Swedish experiments reported on by 1. Holmstrom in 1977. In general, the lime mortars con- taining a higher proportion of brick dust were stronger than those con: taining a lower proportion, For three of the brick dusts, strength appears to increase with reduction in particle size. This trend is more apparent for the higher brick dust o coarse stuff ratio (Fig. 10: samples A6, A18, A24) ‘As seen at Hadrian’s Wall, the ‘compressive strengths of the mortars with small quantities of white ordi- nary portland cement or sulphate resisting cement (Fig. 10: samples B3, BA, B7) were generally lower than those of lime mortars with brick dust. This was the case until a ratio of 1:12 (cement.coarse stuff) or approximately 1/4:1:3 (cemert:lime: sand) was reached. ‘The lime:vement:brick dust mor- tars (C group) were all stronger than the mortars of 1:12 cement:coarse stuff mean. N/mm’ 0.00 Fig, 10. Results: Campressive strength tests Water vapor permeability (WVP). ‘As would have been expected, the WYP of the lime mortars with white ordinary portland cement or salphate- resisting cement seem to decrease “with increasing cement content ‘The mortars with the highest white ordinary portland cement or sulphate-resisting cement content {samples BI and BS) had lower WWP than all the lime mortars with brick dust additives (Table 4). ‘Among the lime:brick dust mor- tars, there s some indication that ‘WYP decreases with increasing pro- portion of brick dust. For three of the brick dusts, this trend seems to be more marked with decreasing particle size (Fig. 11). ‘This may again he an indication that the larger-size brick dusts act more as porous particulates in the mix than as pozzolanic components, a factor which willbe discussed later in this paper. Sodium sulphate crystallization test. There is some indication that this test may be too aggressive to dis- cover very subtle differences between the durability of different mortar mixes. A large variation in weight loss made ic difficult ro identify trends conclusively in the limessand: brick dust mortars (Appendix C) However, there does seem to be an indication thatthe lowerfired brick lasts [Samples A1-A12) performed better than those fired at higher tem- peratures. This corroborates infor mation from a much earlier study on limespozzolan mortars carried out by the BRE in 1940.? In general, though, this test again confirmed the trends seen at Hadrian's Wall regarding lime:sand ‘mortars containing small additions ‘of cement. Allof the lime mortars with cement:coarse stuff ratios between 1:12 and 1:60 (1/4 to 1/20 part of cement in a 1:3 limessand mix) failed by the end of the 15th cycle (Table 5). The addition of Ibrick dust to the lime mortarcement mixture does not seem to have pro- duced much improvement in the result of this test. “These results are most clear in the bar graph (Fig. 12) where it can be observed that: 1, All of the limesand:brick dust mortars (Group A) survived all cycles. 2. The B mortars containing small quantities of cement (B2, B3, BA, B7) all failed by the end of the 15eh ere. 3. The C mortars containing Jime:sandi:brick dusticement also failed or suffered significant weight loss. Conclusions: Phase | The work carried out in Phase I of the Smeaton Project has allowed us to confirm various trends observed in the field and to refine the variables to be tested and the testing procedures for Phase Il, In general, the prelimi- nary conclusions ean be summarized 2s follows: 1. The addition of brick dust does significantly alter the properties of Tine mortars. This more pro nounced in lime: sands brick dust mixtures of at least 1:3:1. Table 3. Sample No. ali a2 asil eit a7. sa avi a1 allt an aly 1/1 als/l agit avi agi a1v/1 20/1 anvil avi 221 ana bin bat ba bait bs. ba cut an eat Compressive Strength 100mm Cubes Cube 065 090 065 18s 030 240 095 1.50. 095 135 085 110 030 has 070 Cabe2 085 030, 065 138, 090 2.10 030 140 07s 150 030 1s 190 175 ars 155 oss 240 050 080 Oss: 030, 09s 275 Tos 170 ous 030 835 0.50 320 3.60 220 ‘Somted on Descencing Strength Mean Sample N/mm? No. oss but 0.90 asi 06s an 18s cui 090 a4ft 225 on 033 6/1 1s ats os al 143 ate os bon 113 alan 095 asi 1.60) Alor 073 a7 183 a2 1B a2 218 a13/l oss a7 063 a23il 0.83, 2 1.00 asl 90 aun 243 39/1 10 s21/1 173 alsil 05, av 030 aul 8.08 320 oss brn 3.13, a1 345 bart 230 bal 2. Lowrfired brick dusts seem: to have the most positive effect on the strength and durability of lime:sand: bbrick dust mixtures. This trend is ‘more pronounced if the brick dust is a more significart portion of the ‘mix. At this point in the research, however, it not possible to say whether this result is due stricdy to firing temperature or is also related toclay type 3. The addition of small quanti- ties of cement to ime:sand mortars Ihasa negative effect on th strength and durability of the mortars. In ‘general, mortar mixes composed of Mean N/mm? 8.10 or 345 3 243 230 225 2a 135 183 133 180 14s 14s Las 113 1.00 0.95 0.95 0.90 0.90 0.90 oe 0s 0.83 0.73, 0.65 0.65 0.63, 0.58 os 0s 030 lime: sand: brick dust show better strengib and durability than lame: sand:cement mortars until the amount of cement in the mix is ‘equal 10 1/2 the volwne of lime (in a 133 lime:sand mix). It seems from these results and from trends perceived by other researchers both in Europe'® and North America! that che addition of rick dust vo limesand morcacs may improve their performance in ‘wo ways, In the lower size particle range («75 microns or < 38 microns according to some researchers), the ‘ata THE SMEATON PROZECT 30 ee Fig. 11, Results: Water vapor permeability potted against rype ana properton ot bck ist « 42 apraULLeTiN problems and inconsistencies created by cutting large prisms in Phase Compaction has becn more regular ‘Teble 4. Water Vapor Permeability Sorted by increasing germeabilty Sample Water Vapor Sample Water and more carefully controlled. In No. Permeabily Now Permeability addition, certain tests have been per- w100mm'/day w/100mm/day formed on samples which were byt 039 av an formed between pairs of saurated but asi a7 bricks. ain an on ‘The tests for moisture content, aot cat a7 stiffening rate, compressive strength saat an 74 and water vapor permeability all aan on oe gave useful information and are an an 078 being repeated in Phase i. ait Son 080 ‘Additional tests have been made aint ban aso con the fresh mortar for consistency aust aun oa using a flow table and on the hard x x ened mortar for porosity since recent an He a research has indicated important cor~ boi aah 096 relations between prosity/parosime- 23/1 bat 098 try and the strength and durability of afl mmortars."4 ‘The sale crystallization test has been altered to be less aggressive, in an attempt to reveal more subtle dif- ferences berween samples.15 brick dust acts as pozzolan, which which seemto have improved perfor- Data has now been collected for reacts with the lime to speed setting mance in Phase _most of the samples tested in Phase and create @ higher strength Based on the experience of Phase II. Complete test results are expected hydraulic mortar, 1, the prepacation of samples was by July, 1994, In the higher particle size renge _altered to afford better control ané In all cases, laboratory results for {> 300 microns) the brick dust tends tomore closely simulate field prac- durability will be correlated in the toact more asa porcus particulate tice. Specifically, groups of smaller Jong term with results observed on orair-entraining additive. The air moulds have heen used to create the samples placed on the exposure content of the aggregate aids carbon- samples foreach test to avoidthe racks. ation and improves resistance to frost and salt crystallization. f In the former case, it seems that a firing temperature below 950°C pro- 100 duces the best quality brick dust for | yap addition to lime mortars. Inthe later case, the porosity of | #190 the brick dust may be more signifi- cant than fising temperature." Phase i ‘These hypotheses are being tested in Phase Il of the Smeaton Project in limited series of mixes utilizing low- fired brick dusts with particles in the ‘exiremes of the particle size range. A ‘non-pozzolanic porous particulate is also being tested to attempt to distin- guish berween the two phenomena Fig 12. Rsuts: Sat crystallzation tests, ‘Table 5. Salt Crystallization Test Sorted by increasing weightloss ‘Sample Mein Ne. % Weight Loss srt 9a? bin 13.43, ail 1642 av 1737 asi 1802 an, 19.85 asi 2083, al 2113, ail 2332 alot 2347 au 2529 ait 2621 alsit 2676 alait 29.32 au 3007 a3! 30.9 ava 32.14 Future Phases Future phases ofthe project will attempt to provide a more in-depth understanding and more conclusive statistical data regarding the material parameters which affect the behavior ‘of pozzolanic additives in lime mor- tars, especially 2s regards particle size, mix ratio, and clay type (in the case of brick dusts). ‘The project will also begin to investigate hydraulic limes and to compare the behavior of hydraulic lime mortars with those previously studied. This isan important area of concern, especialy in view of emerging standardization in the European marketplace. Trials will attempt to characterize and evaluate the performance of various hydraulic limes. OF equal concern, however, are practice parameters, which are, per haps, more difficult © quantify but of great importance in the perfor- mance of mortars, Final phases of the project will atempe to quantify performance variables of fresh and hardened lime- based mortars in respect of issues such as: Sample Mean No. % Weight Loss 2 3321 aan 3338 6) 344 agit 34380 23/1 35.28 s19/1 3722 201 3930 201 4135 18/1 4747 et 7439 ait 9. at f2 wart 42 ont £13

You might also like