Remediation and Protection of Masonry Structures with Crystallising Moisture Blocking Treatment

Abstract

The use of crystallisation materials, so-called moisture blockers, presents an alternative approach to remediating water ingress in masonry structures against conventional time consuming and costly injection of resins. The central claim is that they use the available water in the masonry to block the passage of water to the surface of the masonry, filling pores, cracks and spaces at the interface between mortar and bricks. The aim of this project was to test the effectiveness of new and innovative moisture blocking products at controlling water ingress through masonry. This was done by constructing masonry wells with lime and cement based mortar that were to hold water, and then applying the moisture blocking products at dry and wet condition to the negative hydrostatic pressure side. The moisture levels of both the surfaces and the substrate were then observed for 14 days. The result demonstrated that moisture blocking is an effective method at reducing the levels of surface moisture for bricks, mortar-brick interface and mortar.

Key words: Masonry arch structures, Water ingress, moisture blocker, surface moisture, lime mortar, cement mortar, brick

123456789

101112131415161718192021222324252627

BACKGROUND

The UK transportation network operates some 40,000 masonry highways and 33,000 railway arch bridges. Many of these structures are over 100 years old and are subject to increased traffic loads, speeds, and progressive weathering resulting in structural deterioration. Water ingress in such structures is a widespread problem, traditionally addressed by time consuming and costly injection of resins, with unpredictable success. Water ingress problem areas in masonry are often deliberately obscured by the introduction of canopies and sheeting, possibly with integral water collection. For aesthetic reasons, and because of the possibility of concealing destructive deterioration, this tends to be unacceptable, especially in the case of heritage structures.

Portland cement was not invented until 1828, thus masonry structures prior to the mid nineteenth century were constructed using so called natural mortars, with lime as the binding constituent. Masonry structures later than this generally include Portland cement as the binding constituent (Grimm, 1982; McKee, 1973). Under the action of water ingress, there is a tendency for the lime component to be washed out in the case of natural cement masonry, potentially causing hidden voids within the layers of masonry. This is often referred to as ‘lime leaching’. However, high porosity and other small-scale water transmitting defects can be a problem with both types of mortar.

In mortar, there are two types of pores, interlayer space or very small capillary pores and larger capillary pores that contribute to permeability within a hydrated mortar mix. The former is approximately 0.015µm in diameter and account for 28% of the hydrated mortar volume. The larger capillary pores have a diameter of approximately 5µm and constitute 0-40% of the hydrated mortar volume (Ghosh and Melander, 1991; Luping and Nilsson, 1992; Mehta and Manmohan, 1988). Neville (2011) reported that pores larger than 5µm are entrapped air and are unrelated to permeability. Cement fineness has been reported to influence water ingress through mortar (Powers et al., 1954; Nyame and Illston, 1981). However, researchers reported that that the factors central to water ingress are the water-cement ratio and curing conditions (McMillan and Lyse, 1930; Wiley and Coulson, 1938).

Bricks that create masonry structures are formed mainly from clay (Hendry, 2001). This type of bricks may reach a compressive strength of 100 N/mm2, although bricks with an average of 30 N/mm2 are enough for the construction of houses and walls. Another constituent that takes part in the formation of masonry structures is mortar (Hendry, 2001). Mortars contributes in less than 10% of the total volume of the masonry structure, however it has an enormous effect on the stability and durability of the structure.

Moisture when exists, and it does exist, in bricks it is considered the main reason for the deterioration of masonry structures, and when different types of salts coexist in these structures the condition gets worse (Franzen and Mirwald, 2009). D'Agostino (2013) reported that the process in which moisture contributes in the deterioration of these structures is hard to be fully cognized, as there are multiple mechanical and chemical activities weigh in its movement. It was found that a brick material would take longer to become fully saturated if it had a small threshold diameter (maximum diameter of continuous pores) as this exhibited a lower permeability. Also, it has been found that differences in pore size between the interior and exterior parts of the brick will contribute to brick deterioration. With a surface that has a low porosity obstructing the flow of water through the brick, and a centre that has a high porosity, the centre can become saturated which can lead to problems of freeze-thaw deterioration (Hansen and Kung, 1988; McBurney, 1929). Spalling brickwork is a problem that is a direct result of this phenomenon, water ‘entrapped’ behind the face of the brick can freeze and expand which in turn will mechanically deface the brick. Continual cycles of wetting and drying may also contribute to gradual deterioration of masonry as it causes contraction and expansion in the bricks and may result in cracks at the brick/mortar interface. This leads to an increase in the number of entry points for water ingress to take place.

Newman and Whiteside (1981) conducted a study in which it was found that just 17% of total leakage was due to water ingress at the brick and mortar alone, which suggests that a large amount had transpired at the brick-mortar interface. The bond strength between the brick and mortar is undoubtedly a highly influencing factor of crack initiation and development at the brick-mortar interface. The surface texture of the brick and mortar consistency also play a part in bond integrity; however, the main influence of bond strength is the water retention capacity of the mortar, and rate of absorption of the brick (Palmer, 1931).

28293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485

A research made by Pazderka and Zigler (2013) on protecting heritage masonry buildings, advised to use some systems like bitumen and plastic coatings to protect masonry structures from water attacks, as they found that these systems offer the most effective solutions in terms of sustainability.

Many types of protection materials do exist and they are widely used for protecting concrete (Christodoulou, et al., 2014; Bubalo, et al., 2014; Al-Kheetan, et al., 2017b; Al-Kheetan, et al., 2017c) and, on a narrower level, protecting masonry structures. Cementitious coating materials have been tested to treat water tanks and swimming pools, and they have shown some promising results in terms of protecting the structures from deterioration and in terms of keeping the quality of water within very good limits (Al-Kheetan, et al., 2017a). However, this kind of materials were only tested on concrete structures without considering masonry structures. Protection of porous material by means of impregnation, like dual crystalline technology, is increasingly gaining popularity in civil engineering industry as it improves waterproofing and durability of porous media by filling voids with a non-soluble, highly resistant crystalline formation (Perkins, 2002). A schematic conceptual demonstration is given in Figure 1 to demonstrate how dual crystalline material interact with water. The combined hydrophilic and hydroscopic actions creates moisture blocking mechanism to protect surface against water ingress.

Figure 1: The function and interaction of Dual crystalline materials

The use of crystallisation materials, so-called moisture blockers applied to the negative water pressure side of masonry, presents an alternative approach to remediating water ingress in masonry. Their possible use is an attractive prospect because drilling is not required and the appearance of the masonry unaffected following treatment. The central claim is that they use the available water in the masonry to block the passage of water to the surface of the masonry, filling pores, cracks and spaces at the interface between mortar and bricks.

RESEARCH OBJECTIVE AND SCOPE

The aim of this research is to develop an understanding on moisture blocking techniques in reducing water leaking in the masonry structures. A quantitative and qualitative analysis is provided on the effectiveness and differences of two moisture blocking materials on cement and lime mortars, as well as on clay bricks. This approach will deliver a wider understanding of how water-based moisture blockers work and the scenarios in which they are best applied. The pursuit of possible environmentally friendly and sustainable materials for use in the construction industry is the key driver of this research.

Two water retaining wells are constructed; made from clay brick, one using a lime based mortar and the other a cement based mortar. Two different moisture blocking materials are used, these separately applied to two different sides of the water wells, with one side of each water well left untreated as a control surface. The moisture blockers are applied to the negative hydrostatic pressure side i.e. the exterior of the water wells, for the determination of their ability to dry the exterior surface. The performance of the moisture

8687888990919293949596979899

100101102103

104105106107108109110111112113114115116117118119120121122123124125126127128129130

blockers when applied on brickwork with cement mortar and brickwork with lime mortar are investigated. The remaining side of each water well have a moisture blocking material applied while the surface is dry. This allows a comparison of the effectiveness of the moisture blockers when applied to initially wet or initially dry substrates.EXPERIMENTATION

SpecificationTo determine the effectiveness of moisture blockers, two water wells were constructed (as shown in Figure 2) the dimensions of which are 550mm x 550mm x 675mm. They are constructed using the following materials:

(a) - Clay bricks bonded with Natural Hydraulic Lime (NHL) based mortar (b) - Clay bricks bonded with Portland cement based mortar

Class B bricks were used in this test. The NHL mortar mix consisted of grade 3.5 NHL, which complies with the related British Standard BS EN 459-1: 2015, with sharp sand and water ratio of 1:3:0.45. The cement mortar consisted of Portland cement, sand and water in the ratio 1:3:0.45. These ratios were measured by volume as specified in Eurocode 6 (BS EN 1996-2, 2006; BS EN 1996-3, 2006; Roberts and Brooker, 2014). Great care was taken when mixing the mortars as to achieve the best workmanship level in the water wells. A conventional cement mixer was used and to achieve a consistent and thoroughly mixed mortar, each batch of mortar was mixed for 10 minutes in the mixer.

A moisture tolerant copolymer sealant was used to seal any visible defects in the water well, for example gaps in the brick/mortar interface. The water wells were cured for a period of two weeks to allow time for the mortars to properly harden and form bonds with the bricks. This is important as majority of water ingress in masonry occurs at the brick/mortar interface, and it was imperative that these water wells set as to achieve a standard of water well that will provide reliable results.

aa

b c

Figure 2: Test water wells

Two different moisture blocking materials were tested; MB1 and MB2. The technical specification for both materials are given in Table 1.

Table 1: Technical data for MB1 and MB2 Property MB1 MB2Specific Gravity 1.10 1.09Viscosity (centipoise) 2.2 2.4Freezing point -5oC -4oCBoiling point 110oC 105oCColour Clear Clear Odour None None Toxicity None None Fumes None None Flammability None None Cleaning Water Water

a b

131132133134135136137138139140141142143144145146147148149150151152153154155156157158

159160161162163164165

Drying time (25oC) 2-3 hrs 2-3 hrs

The sequence of material application is shown in Figure 3. MB1 was tested on two sides of the water wells under wet and dry application conditions. MB2 was tested on two water well sides under wet surface conditions. One side of each water well was left untreated as a control surface.

Figure 3: Plan view of water wells showing treatment of each external face

Both depth and surface moisture readings were taken over the course of 14 days, at 3 different heights on all sides of the water wells, 230mm, 380mm and 530mm from the base of the water well as shown in Figure 4. These are referred to as the low, middle and high points respectively in this paper. These points were free from efflorescence. It should be noted that the presence of some efflorescence on the surface does not have any impact on the moisture reading.

Figure 4: Diagram illustrating points where the readings by MMS2 were taken

A Protimeter MMS2 moisture meter, with surface and internal moisture reading capabilities was employed. With this, surface moisture readings use the electrical resistance value between two pin electrodes that are placed on the test surface. Buried moisture is determined according to the dielectric constant measured by radio frequency waves projected into the solid. Readings are measured as percentage wood moisture equivalent (%WME) and relative wood moisture equivalent (RWME) respectively. WME is the moisture level in any material as if it were in close contact and in moisture equilibrium with wood expressed as a % moisture content of wood. %WME values below 17 are deemed dry, between 17 and 20 are at risk, and anything above 20 is wet. RWME values below 170 are deemed to be dry, values between 170 and 200 are deemed to be at risk and finally values above 200 are deemed to be wet.

UV dye was added to the well water with the intention of making leaking water more easily detectable on well surfaces. A portable illuminated digital microscope was employed to capture image of the well surfaces.

Experimental procedures

MB2-surface

wetNT NT

MB1-Surface Wet

MB2-Surface

Wet

MB1-Surface Dry

MB1-Surface Wet

MB1 –surface dry

Bricks with Lime

Mortar

Bricks with Cement Mortar

166167168169170171172173174175176177178179180181182183184185186187188189190191

192193194195196197198199200201202203204205206207208209

Firstly, the initial surface and depth moisture readings were obtained at the three different heights on the exterior faces of each well. MB1 and MB2 were then applied to the allocated dry condition sides. On day 1, the wells were filled with water and the UV dye added to the contained water. The UV torch was then used to detect any major leakages. These leakages were then sealed with waterproof sealant to stop leaking. On day 2, the moisture readings were taken at the same three points on each side of each well. Once the wells had been refilled on day 2, MB1 and MB2 were applied to the wet sides, in the same fashion as before. On day 3 depth and surface moisture readings were then taken at the same three points on each face of each well. The wells were then refilled with water. The process was repeated until day 14, when the final readings were taken.

RESULTS AND ANALYSIS

Crystallisation of moisture blocker

Image of moisture blocker crystallisation were captured using a digital microscope, viewing the brick and mortar surfaces at approximately X550 magnification. This was carried out on every face of each well, including the faces that were left untreated with the aim of obtaining a point of reference. The images captured at the end of the experiment are shown in Figure 5. Irrespective of the type of moisture blocker, crystals are formed on most of the treated surface. The crystals are smooth and transparent on the low textured surfaces of cement mortar and smooth faced, with greater build on lime mortar and coarser face bricks. It appears that the surface texture of the substrate plays influences on the development of crystals on the surface.

Untreated smooth brick face

Untreated rough brick face (some small salt deposits)

Untreated lime mortar Untreated cement mortar

Treated with MB1-dry smooth brick face (subtle transparent surface crystal

Treated with MB1-dry rough brick face (surface crystals forme)

Treated lime mortar with MB1-dry (some crystallisation in larger voids)

Treated cement mortar with MB1-dry (transparent surface crystals)

Treated with MB1-Wet smooth brick face (transparent surface crystals)

Treated with MB1-Wet rough brick face (surface crystals formed)

Treated lime mortar with MB1-Wet (surface crystallisation occupying larger void)

Treated cement mortar with MB1-Wet (transparent surface crystals)

210211212213214215216217218219220221222223224225226227228229230231232

Treated with MB2-Wet smooth brick face (crystals undetectable visually)

Treated with MB2-Wet rough brick face (crystals undetectable visually)

Treated lime mortar with MB2-Wet (surface crystallisation visible)

Treated cement mortar with MB2-Wet (crystals undetectable visually)

Figure 5: Microscopic image of surfaces

At-depth Moisture at Brick-Mortar InterfaceThe depth moisture readings for both brick-lime base mortar and brick-cement based mortar interface are shown in Figures 6a-4f.

0100

200300400

500600700

0 5 10 15

Woo

d M

oist

ure

Equ

ivel

ant

Time (days)

untreated MB2 - Wet MB1 - Dry MB1 - Wet

0100200

300400500

600700

0 5 10 15Woo

d M

oist

ure

Equ

ivel

ant

Time (days)

untreated MB2 - Wet MB1 - Dry MB1 - Wet

Figure 6a: Depth moisture readings at high point-lime mortar and brick interface

Figure 6b: Depth moisture readings at high point-cement mortar and brick interface

0100200

300400500600

700

0 5 10 15Woo

d M

oist

ure

Equ

ivel

ant

Time (days)

untreated MB2 - Wet MB1 - Dry MB1 - Wet

0100200300400500600700

0 5 10 15

Woo

d M

oist

ure

Equ

ivel

ant

Time (days)

untreated MB2 - Wet MB1 - Dry MB1 - Wet

Figure 6c: Depth moisture readings at middle Point-lime mortar and brick interface

Figure 6d: Depth moisture readings at middle point-cement mortar and brick interface

0100200300400500600700

0 5 10 15Woo

d M

oist

ure

Equ

ivel

ant

Time (days)

untreated MB2 - Wet MB1 - Dry MB1 - Wet

0100200

300400500

600700

0 5 10 15Woo

d M

oist

ure

Equ

ivel

ant

Time (days)

untreated MB2 - Wet MB1 - Dry MB1 - Wet

Figure 6e: Depth moisture readings at low point-lime mortar and brick interface

Figure 6f: Depth moisture readings at low point-cement mortar and brick interface

233234235236237238239240

As expected, the level of surface moisture increases with depth under the action of increased water pressure with depth. The results are analyzed according to the depth reading, are discussed in the following.

High sampling pointAt the high sample point (Figures 6a-6b), each face of the well starts at similar levels of moisture in the range of 180-230 relative wood moisture equivalent (RWME). The level is marginally lower with cement based mortar than the lime mortar. By day 4, the untreated face of the well starts to show increased readings of depth moisture at a relatively fast rate, however, this starts to plateau around day 12, at around 340 RWME. The treated sides’ moisture levels also increase by day 4, albeit at a much slower rate and, by day 6, started to steadily declining, presumably marking the point in which the moisture blockers take effect. Interestingly, by day 12 the depth moisture levels on the treated sides of the well have decreased beyond the initial levels of moisture, and then continue to decrease. The face that was treated with MB1 while dry has depth moisture levels that do not reach quite as much as the other faces, but at this height the differences are marginal.

Unlike lime mortar, the moisture level for cement based mortar increases from the outset, but settles by day 3 to similar moisture levels to those recorded for the lime mortar (about 210-230 RWME). The relative dryness of the cement mortar before the wells were filled is likely to be the cause for this, absorbing moisture relatively fast at first. Like the lime mortar, the treated faces of the cement mortar well start to steadily decrease in depth moisture by days 3 and 4, with no significant performance difference between the two moisture blocker materials or their or application conditions. It is from this point that the untreated face steadily increases in at-depth moisture levels for the remainder of the experiment. The marginal discrepancies in the moisture level readings taken from day 3-4 are probably due to varying degrees of workmanship. The influence of this appears to be a decrease as the moisture blockers starts to initiate and crystallizing the surface.

Middle and low sampling pointsThe middle (Figures 6c and 6d) and lower sampling points of the NHL well (Figures 6e and 6f) behave similarly in terms of at-depth moisture levels, all of which start at just over 200 RWME. The wet treated and untreated sides of the well have a relatively rapid increase in depth moisture levels compared with the high sampling points, reaching their peak values by day 2-3. At the middle sample point, the two wet treated sides peaked at around 420 RWME with MB1-dry peaking at 377 RWME, whilst at the low sample point the wet treated sides and MB1-dry peaked at around 500 and 420 RWME respectively. This increase in moisture is more contrasting with the lower sample point relative to the middle sample point; the reason being twofold:

• There is larger hydrostatic pressure acts at the base of the well, driving greater moisture transmission.

• Water levels in the well dropped due to leakage between refilling, causing the water pressure to periodically reduce to the top sampling point.

By day 4, depth moisture levels of the wet treated sides start to decline, steadily toward approximately 250 RWME at the end of the experiment. Among the wet treated sides, MB1 seems to have a faster influence on at-depth moisture levels, however, by day 7 at the middle sample point and day 5 at the lower sample point, the RWME values are very similar. The at-depth moisture levels of the untreated face show readings that follow the trend of the other sides until day 3, where levels continue to increase to approximately 500 RWME at the middle sample point and 600 RWME at the lower sample point.

Side MB1-dry shows an increase in at-depth moisture that is less pronounced, with a significantly lower peak moisture level than the other sides. This is probably attributable to the opportunity that the masonry could absorb the moisture blocker into the substrate before the water ingress had commenced, allowing it to initially activate deeper into the core of the structure than the wet treated sides. This would also explain why the surface crystallization on the dry treated surface was much subtler than on the wet sides (explained in earlier section). By day 4-5 the moisture levels on the MB1-dry side at the lower and middle sample points respectively, have peaked and converged with the other sides. Levels of moisture in all treated sides steadily decline at similar rates once converged, with no side significantly drier than another.

241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298

In terms of mortar type, the at-depth moisture levels in the cement mortar behaved very similarly to the lime well, wet treated sides incurring a faster rise and peak in moisture levels than the dry treated side. The untreated sides increase as the treated sides, before settling and plateauing at the end of the experiment. There is a small but significant spike in the moisture levels with MB2-wet and MB1-dry sides at the middle point on day 4 (Figures 6c and 6d). This is probably attributable to small leaks at the brick mortar interface.

Treatment of initially wet vs dry substrateThe at-depth moisture readings at the lower sample point peak on day 2 at 500 RWME on the wet treated faces, whereas MB1-dry moisture levels peak from day 5-7 at 448 RWME. The readings for each treated face, as with the NHL well, converge around day 7, with MB1-dry levels marginally but consistently smaller than the other two treated faces. The at depth moisture readings in the first 4 days of the dry treated face of both wells show a trend, the levels of moisture increasing from day 1-2 and then settling, before ageing increasing. It is probable that water initially travels through the pores in the substrate where it reaches the moisture blocker, where it then activates the hydrophilic crystals leading to the settlement of the levels. The water then finds a route around this hydrophilic crystallization through the brick/mortar interface leading to a secondary rise in moisture levels. Crystallization then starts to form at the interface causing the moisture levels to reduce and the substrate to dry. The depth readings at the middle and lower sample points of both wells did not decrease all the way to the starting levels, however the steady and consistent decline of readings suggest that given more time there is a large possibility that they would have.

Effectiveness of moisture blockersThe lower sample point data is used in Table 2 to illustrate the difference in moisture levels between the two wells and their faces, as it is here that the most contrasting effect of the moisture blockers occurred. The at-depth moisture levels of the NHL mortar were significantly higher than those of the cement mortar prior to the filling the wells, with the difference in RWME in the range of 15-27 between the corresponding external faces of each well.

Table 2. At-depth moisture levels at the low sample point of each face of wells prior to filling, peak and final readings.

Depth Readings (relative WME)

Brick/Cement Brick/Lime

Face of Well NT MB2 Wet

MB1 Wet

MB1 Dry

NT MB2 Wet

MB1 Wet

MB1 Dry

Prior To Filling 187 185 181 177 202 208 208 198Peak Moisture 592 504 499 431 597 510 504 454End Moisture (day 14) 592 369 356 344 597 339 342 336

Interestingly, the two water wells have very similar peak moisture values at each face with differences in RWME of as little as 6. By the end of the experiment, the NHL mortar depth moisture levels have even reduced below that of the cement mortar (albeit marginally). The difference in completion moisture levels between all treated faces of both wells, was even small with the largest and smallest values of 369 and 333 RWME respectively. Should the experiment have continued, there is a possibility that this difference in moisture levels at this depth may have increased, as the ability of lime mortar to allow water vapour transmission, allowing faster drying of the substrate to transpire MB1-dry has a lower peak at-depth moisture level. However, by the end of the experiment it was not significantly any drier than the wet treated sides.

Surface Moisture for Mortar The WME (%) profiles at high, medium and low sample points are similar for the mortar-brick interface, although the magnitudes are significantly lower. Therefore, for brevity, only key measurements are shown in Tables 3 and 4.

Table 3. The surface moisture levels of the lime mortar at each face of wells prior to filling, peak and final readings.

Surface Readings (%WME)

Lime mortarHigh sample point Medium sample point Low sample point

NT MB1dry

MB1 wet

MB2wet

NT MB1dry

MB1 wet

MB2wet

NT MB1dry

MB1 wet

MB2wet

299300301302303304305306307308309310311312313314315316317318319320321322323324325326327

328329330331332333334335336337338339340341342343344

Prior to filling 31.9 36.7 34.3 36.3 32.3 38.6 34.1 34.9 39.3 38.9 36.6 34.3

Peak Moisture 56.6 46.1 47.5 47.1 65.1 57.3 55.1 56.1 73.7 68.7 68.8 64.6

End Moisture 56.6 39.7 43.6 42.3 64.7 51.1 48.4 48.8 73.7 55.6 52.1 54.2

Table 4. The surface moisture levels of the cement mortar at each face of wells prior to filling, peak and final readings.

Surface Readings (%WME)

Cement based mortarHigh sample point Medium sample point Low sample point

NT MB1dry

MB1 wet

MB2wet

NT MB1dry

MB1 wet

MB2wet

NT MB1dry

MB1 wet

MB2wet

Prior to filling 17.1 15.6 22.4 18.1 20.1 18.2 23.2 21.3 18.3 20.5 25.4 22.9

Peak Moisture 51.1 43 43.4 45.8 53.9 54.5 46.4 54.3 64.8 54.9 51.2 53.5

End Moisture 51.1 41.9 43.4 45.6 53.9 53.8 46.4 50.8 64.3 46.1 50.1 42.9

Surface moisture at high, medium and low sample pointsAt the high sample point, the surface moisture on each face of the lime mortar started at similar levels in the range of 31-37 %WME, although the rate of increase for each face was found to be different. The surface moisture of side MB2-wet peaks by day 4 at 47.1%, whilst sides MB1 (wet & dry) peak on day 5 at 47.5% and 46.1%WME respectively. The untreated side, as expected, continues to increase throughout the experiment. The sides MB1-dry and MB2-wet decrease in surface moisture from their peak in a similar rate and fashion.

The surface moisture of the cement mortar well at the highest sample point was like that of the lime mortar, but there are some differences in moisture levels between the two. The MB1-dry and MB2 surface moisture increase at similar rates to each other, initially faster than the untreated and MB1-wet sides. The moisture levels of the different sides converge and are similar by day 5. MB1-wet & MB2-wet moisture levels peaked by day 6 at 42.1 & 45.8%WME respectively, before declining in the following days. This declination increases by the 9th day of the experiment and the surface drying rate increases, with MB1 consistently dryer than MB2.

Both the middle and lower sample point moisture readings for the lime mortar start at similar readings about the range of 32-40%WME, where they all increase at similar rates apart from MB1-dry that is slightly slower. As expected, the lower sample point had a higher rate of moisture level increase. The untreated surface moisture continues to increase and plateaus at about 65% and 73%WME at the middle and lower sample points respectively.

The mortar surface moisture values of all treated sides reached a peak by days 5-6, at the middle height sample point, with a %WME range of between 55-58%. At lower sample points peaked at days 4-5 with a range 64-68%. At the lower sample point MB1-dry has a slower initial rate of surface moisture build up, although it peaks slightly higher than the other treated sides. This initial rate is probably due to the activated moisture blocker material in the substrate, holding off some of the water ingress. The secondary rise is likely to be caused by water that has travelled around this crystallisation through the brick/mortar interface.

Initially wet vs dry treated substrateAt both sample points on the lime mortar, MB1 is slightly more effective at drying the surface of the mortar than MB2, having said this it would remain to be seen whether this difference would be significant in long term real world scenarios. The same can be said when comparing MB1’s effectiveness when applied on wet or dry conditions, MB1-dry appears to have been marginally more effective but the significance of this small difference is not determined.

345346347348349

350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386

At the lower sample point, the wet treated sides and NT side have a fast-initial increase in mortar surface moisture, before slowing down, the wet treated sides continued to increase until day 5 where they peaked at 51-54%WME. Side NT continued to rise a plateaued at around 64%WME. MB1-dry moisture levels on the other hand increase slower than the other sides at first, but is a steady increase and peaks at 55%WME on day 5, along with the other treated sides. By day 6, moisture levels on the mortar for all treated sides were declining, and by day 9 of the experiment the declination of moisture levels at all sides had strengthened. This increase in declination of mortar surface moisture has been observed at every sample point of both wells, except the sample points where some leakages took place. This is due to seepages at the brick/mortar interface; the seepage inhibiting the surface drying properly of the blocker, until enough crystals are formed at the interface (through the hygroscopic and hydrophilic nature of the material). This then reduces seepages and allows faster drying of the masonry surface.

Lime vs cement mortarThe surface moisture of lime mortar starts higher than cement and stays consistently higher throughout the experiment, even more so at peak moisture with the difference decreased by the end to testing. For example, the moisture difference between the two wells on their MB1-wet sides throughout was:

• Prior to filling – 11.2%WME difference• Peak moisture levels – 17.6%WME difference• End moisture levels – 14%WME difference

This suggests that, given a longer period of monitoring, the surface moisture of the lime mortar might have reduced to that of the cement mortar. The difference in surface moisture at the MB2-wet sides on the other hand increases between the peak – 11.1%WME difference and end – 18.1%WME difference. MB2-applied to the wet substrate seems to have been significantly more effective on reducing surface moisture on cement mortar than lime mortar. It was observed that MB1-dry did not reach depth moisture readings as high as for the other treated sides. This was not the case for the moisture levels on the surface of the mortar. This is because of water travelling through small brick/mortar interface gaps and collecting on the surface, rather than a result of water exclusively traveling through the capillaries of the mortar matrix.

Surface Moisture at BrickThe WME (%) profiles at high, medium and low sample points are like mortar and brick-mortar interface, although the magnitudes are significantly lower. For brevity, only key measurements are shown in Tables 5 and 6. As with the mortar surface moisture readings, surface moisture on bricks show a large spike of moisture on day 13 and maintained at this level of moisture in day 14. This could be due to humidity observed at the end of experiment. It should be noted that although day 14 was taken as final reading, the high surface humidity may have affected the results at this stage.

Table 5. The surface moisture levels on the lime mortar for each face of wells prior to filling, peak and final readings.

Surface Readings (%WME)

Lime based mortarHigh sample point Medium sample point Low sample point

NT MB1dry

MB1 wet

MB2wet

NT MB1dry

MB1 wet

MB2wet

NT MB1dry

MB1 wet

MB2wet

Prior to filling 11.8 15.3 14.7 15.3 12.4 14.8 12.5 16.2 12.3 15.4 8.9 15.4

Peak Moisture 34 33.2 28.4 29.9 38.9 38.6 29.7 34.8 49.6 42.6 41.3 38.9

End Moisture 34 33.1 28.3 27.6 38.9 36.3 29.7 34.8 49.6 42.6 37.1 35.5

Table 6. The surface moisture levels on the cement mortar for each face of wells prior to filling, the peak and final reading.

Surface Readings (%WME)

Cement based mortarHigh sample point Medium sample point Low sample point

NT MB1 MB1 MB2 NT MB1 MB1 MB2 NT MB1 MB1 MB2

387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427

428429430431

dry wet wet dry wet wet dry wet wetPrior to filling 20.9 19.8 16.9 17.5 19.7 20.7 18.5 19.4 20.1 21.9 21.8 19.3

Peak Moisture 37.9 32.3 36.4 34.3 44.9 37.2 43 35.7 51.4 48.4 42.9 40.7

End Moisture 36.2 31.3 35.3 28.3 44.9 37.2 41.8 34.8 51.4 48.4 38 36.1

For lime based well (Table 5), treated sides commence at similar moisture levels of approximately 15%WME, NT and 11.8%. The NT values increases steadily and levels out at 30.5% by day 12. MB1-dry peaks on day 5 at 25.3%WME, earlier than the other treated sides, most likely since the MB material had time to be absorbed into the brick and was crystallising in the substrate, inhibiting water ingress. MB1-wet and MB2-wet surface moisture steadily increased, albeit MB1 at a significantly slower rate and peaked at 21.6% and 26.4% WME respectively on day 6. After peaking, the surface moisture of the bricks on all treated sides then steadily decline, until day 10 where a spike in humidity occurred and thus, condensation on the surface caused a spike in moisture readings (see Figure 6a).

The surface moisture on the untreated side of the cement well (Table 6) commenced at 19.8%WME where it then steadily increased throughout the experiment, and started to settle by day 12 at 36.7%WME. All treated sides had a similar steady increase in surface moisture, with MB1-wet and MB2-wet sides reaching peak moisture on day 5 at 23.1% and 27.7% WME respectively. MB1-dry surface moisture peaked a day later on day 6 at 27.1%WME, possibly due to crystal formation within the substrate, slowing but not stopping the water ingress through the brick capillary structure.

Brick surface moisture at middle and lower sample points show that all sides of the NHL well commenced similar, ranging from 13-16%WME, where moisture then steadily increased (more so at lower point both in rate and steadiness). As expected at both the middle and lower sample heights, the brick surface moisture on the untreated side continues to rise and ends at 37.6 and 48.4%WME. By day 5 at the middle sample height MB1-wet & dry have both peaked at 31.3% and 28.9% respectively. They then slowly started to decrease in surface moisture, while MB1-dry staying consistently dryer. Moisture at the MB2-wet brick surface continued to rise and did not peak until day 8 at 34.6%. At the lower sample point both MB1-wet and MB2-wet reached their peak surface moisture by day 5 at 38.9% and 41.3% WME, where both started to decline. However, the rate of declination is significantly faster in MB1-wet side than MB2-wet.

Lime vs cement mortarAt both sample heights, all surface readings on the cement mortar well (Table 6) commenced at approximately 20%WME. When the surface moisture of the treated sides reached peak value, the untreated sides at the middle and lower sample points continue to increase and by the end of the experiment were 42.1% and 49.8% respectively. The bricks in the cement mortar well actually started significantly wetter than the bricks in the NHL well. This difference decreases largely by the end of the experiment. The biggest change was on the MB1-dry sides of the wells:

• Prior to filling – 12.9%WME difference• Peak moisture levels – 1.6%WME difference • End moisture levels – 1%WME difference

The fact that every side of the cement well had wetter bricks, suggests that the observed trend cannot be solely attributed to the conditions of the bricks pre-construction. This was observed due to NHL mortars ability to allow water vapour transmission. This allows the moisture in the NHL mortar to attempt to diffuse via the mortar, whereas the moisture in the cement mortar is absorbed, in part, by the bricks surrounding it. This would also explain why the cement mortar started dryer than the NHL mortar.

The final surface moisture on the bricks at the lower sample point on the sides treated with MB1 is significantly lower than that of MB2. It is for this reason a case can be made that the MB1 material is more effective at drying out the surface of clay brick.

CONCLUSIONS AND FURTHER WORK

432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483

Key conclusions drawn from this investigation are;

In general, the moisture blocker treated sides of the water wells were effective at reducing the levels of moisture transmission for bricks, mortar-brick interface and mortar. The NHL mortar surface moisture peaks in general a day before cement mortar. This is a result of a greater general porosity of NHL mortar that allows water vapour transmission, more so than cement and thus leading to a quicker impregnation of moisture blocker crystals within the substrate.

Both moisture blocking materials proved similarly effective at drying out the exterior of the water well masonry. However, MB1 did have faster initial influence on reducing moisture levels. When applied to either a wet or dry surface, MB1 is similarly effective at drying out the exterior masonry substrate. However, where possible, application before exposure to water ingress would substantially reduce initial buried moisture presence (by 68 and 50 RWME for the cement and NHL wells respectively).

It appears that MB1 is more effective at drying and preventing NHL surface moisture than MB2. However, the difference is small and in real world applications possible not significant. MB1, when applied in both wet and dry conditions appear to be more effective than MB2 applied in wet conditions, with MB1-dry also marginally out-performing MB1-wet. Again, this difference between the wet and dry applications of MB1 is possibly insignificant in real world applications, over long periods of time. Overall, MB1 is significantly more effective at drying the surface of the clay brick than MB2.

The readings taken from the mortar and brick suggest that the general porosity and permeability of the clay bricks is lower than that of the mortar matrix. This is apparent from the observation that surface moisture on the brick takes 1-2 days longer to reach peak level values (which is smaller than that of the mortar).

To a great extent, the approach to the research is ad hoc, with no relevant standard approach to water well based testing existing. In the on-going investigations, a standard water well format is being determined, with use of multiple water wells that will facilitate statistical analysis of experimental data.

484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522

REFERENCES

(1) Al-Kheetan, M. J., Rahman, M. M., Chamberlain, D. A. (2017a), “Influence of Early Water Exposure on Modified Cementitious Coating”, Construction and Building Materials, Vol. 141, pp. 64-71.

(2) Al-Kheetan, M. J., Rahman, M. M., Chamberlain, D. A. (2017b), “Influence of Crystalline Admixture on Fresh Concrete to Develop Hydrophobicity”, Transportation Research Board 96th Annual Meeting, Jan 8-12, 2017, Washington D.C.

(3) Al-Kheetan, M. J., Rahman, M. M., Chamberlain, D. A. (2017c), “Influence of Hydrophobic Admixture and Curing Agent on Water-Resistant Concrete”, Transportation Research Board 96th Annual Meeting, Jan 8-12, 2017, Washington D.C.

(4) British Standards Institution. BS EN 459-1 (2015), “Building Lime Part 1: Definitions, Specifications and conformity criteria”, British Standards Institution, London.

(5) British Standards Institution. BS EN 1996-2 (2006), “Eurocode 6. Design of masonry structures. Design considerations, selection of materials and execution of masonry”, British Standards Institution, London.

(6) British Standards Institution. BS EN 1996-3 (2006), “Eurocode 6. Design of masonry structures. Simplified calculation methods for unreinforced masonry structures”, British Standards Institution, London.

(7) Bubalo, M.C., Radošević, K., Redovniković, I.R., Halambek, J. and Srček, V.G. (2014), “A brief overview of the potential environmental hazards of ionic liquids”, Ecotoxicology and environmental safety, Vol. 99, pp. 1-12.

(8) Christodoulou, C., Tiplady, H., Goodier, C.I. and Austin, S.A. (2014), “Performance of silane impregnants for the protection of reinforced concrete”, In Grantham, M., Basheer, P. A., Magee, B., and Soutsos, M. (Ed.), Concrete Solutions 2014, Proceedings of Concrete Solutions, the 5th 26 International Conference on Concrete Repair, Boca Racon 2014, pp. 385-392.

(9) D'Agostino, D. (2013), “Moisture dynamics in an historical masonry structure: The Cathedral of Lecce (South Italy)”, Building and Environment, Vol. 63, pp. 122-133.

(10) Franzen, C., and Mirwald, P.W. (2009), “Moisture sorption behaviour of salt mixtures in porous stone”, Chemie der Erde-Geochemistry, Vol. 69, No. 1, pp. 91-98.

(11) Ghosh, S. K. and Melander, J. M. (1991), “Air Content of Mortar and Water Penetration of Masonry Walls”, Portland Cement Association, Skokie. IL 60077

(12) Grimm, C.T. (1982), “Water Permeance of Masonry Walls: A Review of the Literature”, in Borchelt, J.G. (Ed.), Masonry: Materials, Properties, and Performance, ASTM STP 778, American SocietyFor Testing and Materials, Philadelphia, pp. 178-199.

(13) Hansen, W. and Kung, J.H. (1988), “Pore Structure and Frost Durability of Clay Bricks”, Materials and Structures, Vol. 21, No. 126, pp. 442-447.

(14) Hendry, Emeritus A.W. (2001), “Masonry walls: materials and construction”, Construction and Building materials, Vol. 15, No. 8, pp.  323-330.

523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576

(15) Luping, T. and Nilsson, A. (1992), “A Study of the Quantitative Relationship between Permeability and Pore Size Distribution of Hardened Cement Paste”, Cement and Concrete Research. Vol. 22, No. 4, pp. 541-550.

(16) McBurney, J.W. (1929), The Water Absorption and Penetrability of Brick, Proc. of the American Society for Testing Materials, Vol. 29, Part 2, pp. 711-739.

(17) McKee, H. J. (1973), “Introduction to Early American Masonry - Stone, Brick, Mortar, and Plaster”, National Trust for Historic Preservation, Columbia University, pp. 61.

(18) McMillan, F.R. and Lyse, I. (1930), “Some Permeability Studies of Concrete”, Proc. American Concrete Institution, May 1930, pp.101-142.

(19) Mehta, P.K. and Manmohan, D. (1988), “Pore Size Distribution and Permeability of Hardened Cement Pastes”, The International Congress of Chemistry. Paris, Part 7, 1988, pp. 1-5.

(20) Neville, A.M. (2011), Properties of Concrete, Harlow: Pearson Education Limited, 5th ed.

(21) Newman, A.J. and Whiteside, D. (1981), “Water and Air Penetration Through Brick Walls – A theoretical and Experimental Study”, Trans. of the Journal of the British Ceramics Society, Vol. 80, No.1, pp. 27-36.

(22) Nyame, B.K. and Illston, SM. (1981), “Relationship between Permeability and Pore Structure of Hardened Cement Paste”, Magazine of Concrete Research, Vol. 33, No. 116, pp. 139-146.

(23) Palmer, L.A. (1931), “Water Penetration through Brick – Mortar Assemblages”, Journal of the Clats Institute, pp. 19-31.

(24) Pazderka, J., Zigler, R., (2013), “Refurbishment of Moist Building Masonry in terms of Sustainable Building”, Central Europe towards Sustainable Building: Sustainable Building and Refurbishment for Next Generations, Sustainable Refurbishment of Existing Building Stock, June 26-28, 2013, Prague.

(25) Perkins, P.H. (2002), Repair, protection and waterproofing of concrete structures, CRC Press, 3rd ed. London: Madras.

(26) Powers, T.C., Copeland, L.E., Hayes, J.C. and Mann, H.M. (1954), “Permeability of Portland Cement Paste”, Proc. American Concrete Institute, Vol.51, Nov. 1954, pp. 285-298.

(27) Roberts, J. J. and Brooker, O. (2014), “How to Design Masonry Structures Using Eurocode 6”, [e-book] The Concrete Centre. Available at: http://www.eurocode6.org (Accessed: 24 Mar 2014).

(28) Tait, C.J., Khalaf, F.M. and Kermani, A. (2014), “Review of Developments in Water Ingress through Stressed Masonry”, p. 5. Available at: http://www.globalskm.com/ (Accessed: 3 Mar 2014).

(29) Wiley, G and Coulson, D.C. (1938), “A Simple Test for Water Permeability”, Journal of the Americam Concrete Institute, Vol. 34, pp. 65-75.

577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625