1 CHAPTER 5 Integral Waterproofers for Concrete

2 The need for waterproofing admixtureswell-made concrete has low permeability and porosity (typically the coefficient of hydraulic permeability (10-8 – m/sec). In practice concrete structures often allow the passage of water not only through joints and discontinuities, but through the mass concrete itself. Water can penetrate concrete in two ways: when there is a hydrostatic pressure on one side of a concrete mass, water and aggressive agents can travel through any channels which interconnect the two faces of the concrete. water may be absorbed by capillary action and travel through the concrete to a face where it evaporates because the air in contact with the surface is unsaturated.

3 Admixtures that can reduce permeability will be effective in the first situation.However, admixtures that impart water-repelling or damp-proofing properties (hydrophobers) may reduce the effect of the second mechanism whilst being less effective at preventing water passage under a positive hydrostatic head. A common factor in all concretes allowing the passage of water and/or water vapour is the presence of inter-connected voids. Without such voids and their interconnections water or vapour transfer cannot take place.

4 Voidage splits into three main groups:The gel pores that are minute (approximately 2 nm) and independent of (w/c) ratio. Capillary pores that result from excess water being present and, therefore, depend upon w/c ratio. Entrapped voidage reflecting the concretes rheology and mixing method.

5 One way in which the number of channels and capillaries can be reduced is by controlling the w/c ratio. During the hydration some of the capillary pathways become blocked by reaction products consisting of crystalline material and calcium silicate hydrate gel (CSH gel). CSH gel has a very low hydraulic permeability (7 X m/sec). if the w/c ratio is low enough the volume of gel produced will be sufficient to completely block the interconnecting capillaries within the cement paste. For ordinary Portland cement concrete if the w/c ratio is in excess of 0.7 there will not be sufficient gel formed to block the capillaries so resulting in interconnections. (Fig. 5.1).

6

7 Definition of an integral waterprooferIntegral waterproofer: a material (powder, liquid or suspension) that when intimately mixed with the fresh concrete results in reduction in the hydraulic permeability of the cured concrete mass and/or a water-repelling or hydrophobic property being imparted to the set concrete. This definition excludes accelerating admixtures such as calcium chloride since such materials only alter the rate at which some initial permeability is reached and not necessarily its ultimate value.

8 CLASSIFICATION OF WATERPROOFING ADMIXTURESPermeability reducers Water repellents or hydrophobers (iii) Miscellaneous.

9 Members of each group may be reactive or non-reactive.By reactive we mean that when the admixture contacts wet cement chemical interaction occurs resulting in a new product which imparts the waterproofing property. Non-reactive materials will include the group known as 'inert fillers'. Integral waterproofers range from thick, clear or cloudy liquids, to pastes or dry powders. They may be added to the gauging water or to the concrete mix but, as previously stated, materials that have been pre-mixed with the dry cement are excluded.

10 Permeability reducersWe can sub-classify materials within this group, which reduce the hydraulic permeability of concrete, into very fine particulate materials, workability and air-entraining admixtures, accelerators.

11 Very fine particulate materialsGround sand, whiting, bentonite, PFA, diatomaceous earth, limestone, slag and pumice, colloidal silica (including silica fume) and fluorosilicates. It is preferable for the finely divided material to have some degree of pozzolanic reactivity, so densifying the cement gel matrix by the replacement of coarse calcium hydroxide crystals with finer gel-like hydrated calcium silicate products. Particulate materials are of real benefit if the concrete mix is low in cement or deficient in fines. However, in cement-rich mixes the effect could be the reverse since the addition of fine particles could increase the water requirement, leading to a less dense and lower strength concrete.

12 Workability and air-entraining admixturesIncorporating a workability agent or plasticiser reduces the chance of entrapped air voids and the lower water requirement offsets bleed. The lignosulphonate is adsorbed onto the C3A (tricalcium aluminate) early hydration products, probably giving better dispersion of the cement particles. Incorporation of lignosulphonates may give rise to a finer pore system which, whilst causing high capillary absorption, increases resistance both to actual passage of water and hence to a lower permeability. Air-entraining agents act in a similar manner to lignosulphonates by imparting improved workability to the mix and thus allowing less water to be used.

13 Accelerators The function of accelerators as permeability reducers is doubtful. The use of calcium chloride, for instance, may improve early permeability and porosity figures simply by advancing the overall hydration reaction of the C3A and C3S phases but the ultimate permeability remains unchanged.

14 Water repellents or hydrophobersMaterials in this group reduce the passage of water through dry concrete which would normally occur as a result of capillary action and not as a result of an external pressure of water. In principle it is thought that all these materials impart a water-repellent property to the concrete surface as well as lining and, in some cases, blocking the pores. The detailed mechanism is obscure but it has been suggested that the water-repellent action is associated with an electrostatic charge imparted to the walls of the capillaries. Materials comprising this group are, Soaps Butyl stearate Selected petroleum products and natural oils.

15 1. Soaps These are usually metal salts, the sulphonium salts of fatty acids (calcium, sodium or ammonium stearate or oleate as well as stearic acid dispersions). The pore system that develops during the later stages of hydration (greater than 24 hours) is not affected by this precipitate hence the saturated permeability is not reduced. Many of these materials entrain air, due presumably to their surfactant properties, and dosages exceeding 0.2% m/m of cement are not recommended without the likelihood of significant strength loss. On the other hand, improved workability may result from using these soaps, which offsets the formation of cavities and large voids.

16 2. Butyl Stearate The hydrophobic action is similar to that of the soaps, in that the eventual compound resulting in water repellency is calcium stearate. However, butyl stearate hydrolyses only slowly in the alkaline phase of concrete and slowly produces the calcium stearate. As a result very much less air is entrained and strength reductions are not serious. This allows higher additions to be used, which invariably gives improved damp-poroofness. Slow reaction of the butyl stearate also allows better distribution of the admixture throughout the cement gel.

17 3. Selected petroleum productsMineral oils, waxes, cut-back and emulsified asphalts as well as natural oils such as linseed, rapeseed and vegetable oils comprise this group. In the case of the asphalt emulsions the dispersion is broken by the drying out of the concrete, resulting in hydrophobing and some pore blocking. The other compounds are generally regarded as 'inert' hydrophobers acting mainly in a physical way without obvious reaction with the cementitious components. Some strength reduction may result due to entrained air, particularly if emulsifying agents have been used to disperse the hydrocarbon. Certain wax emulsions may improve the damp-proofness of steam-cured concrete by melting 'in situ' and so blocking the pores.

18 Test methods for damp/waterproofers Method ITwo methods are available. These will be mentioned here. a) The first method is that detailed in BS 556: Part 2 : 1972 or a modification of it. Essentially the concrete to be tested is oven-dried and immersed in the dried-out state in water for fixed periods of time ranging from a few minutes to many hours. Water absorption is measured as the weight change at the fixed immersion times.-

19

20 b) The 'Von Test’ is an extension of the simple absorption test which avoids the entrapment of air in the centre of the specimen. It uses a hollow cylindrical specimen (cast in a standard 150 mm X 300 mm mould with a central insert), the wall thickness being 50 mm (Fig. 5.4). The specimen is almost completely immersed in water contained in a covered vessel, and the amount of water permeating into the central cavity after a given time measured by pipette. In the case of very impermeable concrete, the specimen can be weighed as in Method I(a). Again, results are only relative and cannot be related to any absolute property.

21

22 Test methods for damp/waterproofers Method IIa) The ISA Test This method is widely known as the Initial Surface Absorption Test or IS AT and is contained in BS 1881 : Part 5 : 1970. The test measures the rate of flow of water into concrete per unit area after a stated interval from the beginning of the test. A small constant head of water is used and a steady temperature maintained. The concrete is again oven-dried before carrying out the test. The apparatus is shown in Fig. 5.5.* Briefly the cell or cap (of known area) is clamped or fastened to the surface of the concrete and filled with water from the reservoir. The reservoir supplies a constant head of water of about 20 cm. After entrapped air has been bled off through the central connection the tubing is quickly connected to the calibrated capillary and this is flushed to expel all air. At the start of taking a reading, tap (A) is closed and the movement of the meniscus in the capillary tube is then timed between two selected calibration points. Alternatively, the distance moved through a given time is noted. Readings may be taken at time intervals ranging from a few minutes to several hours. ISAT values are quoted in units of ml/sq metre/second.

23

24 b) Figg Hypodermic MethodTo carry out this test, a hole typically 30 mm deep of 5.5 mm diameter is drilled into the concrete surface, taking care to avoid large aggregate particles. Having cleaned out the hole thoroughly, a disc of polyether foam (3 mm thick X 7.5 mm diameter) is inserted to form a plug 10 mm from the bottom, and the cavity above filled with silicone sealant (Fig. 5.6). A hypodermic needle is then pushed through the plug and connected to a water-filled syringe via an adaptor. Air is displaced from the cell in the concrete and the capillary side arm. When the system is filled with water, the syringe is isolated by means of a stop-cock and water allowed to permeate the concrete under an applied head of 100 mm. The time taken (in seconds) for the water meniscus to travel 50 mm along the capillary tube is then recorded.

25

26 c) Modified Figg MethodIn this modification, a larger hole (16 mm diameter X 40 mm deep) is drilled into the concrete, and a silicone rubber plug formed as before. Two hypodermic needles are inserted through the plug, one connected to a water reservoir and syringe (see Fig. 5.7) and the other to a horizontal capillary tube and scale. The system is rapidly flushed through with water, and the rate of flow of water into the concrete measured at stated intervals of time, as in the case of ISAT. In this case results are expressed in terms of time in seconds for the water meniscus to travel a stated distance along the capillary tube, but this can be converted into a volume flow rate per unit area of the injection cell.

27

28 d) 1 ml and 3 ml Test Both methods measure the time required for 1 ml (in the case of hard materials) and 3 ml (for more open textured materials) of water to be absorbed. The method is simple but quite arbitrary.

29 Method III Q (cc/sec) = Kp (dp/dx) (per unit area)This method really measures the unsaturated steady state coefficient Kp where water vapour is the migrating fluid. Similar methods are used for determining the vapour permeability of paint film or membranes. Q (cc/sec) = Kp (dp/dx) (per unit area) Where dp/dx refers to the vapour pressure potential or driving force, Kp is measured using variations of the dry cup test. The technique is shown diagrammatically in Fig. 5.8. Tablet samples of concrete, mortar or cement paste are sealed at the edge with wax or synthetic resin. The concrete is placed in a chamber containing water vapour at known vapour pressure, the sample is allowed to equilibrate, after which it is attached to an aluminium cup containing desiccant such as anhydrous calcium chloride. It is important to maintain constant temperature. Moisture diffuses through the pores of the concrete and is taken up by the desiccant. Weighing the cup and specimen at regular intervals shows when the 'steady state1 has been reached. Plotting values of Q X (X— specimen thickness) against P we obtain a smooth curve represented by the function:

30 By graphical differentiation the unknown coefficient Kp may be obtained. Plotting Kp against vapour pressure often gives a very good semi-quantitative indication of performance of one admixture against another. Unfortunately the method is restricted to small specimens. A rather special case of water transport through a specimen is when we have liquid on one side and vapour transport through the sample and out the other side. Here we have a condition of saturated and unsaturated permeabilities. It is possible by using a dry cup upside down and having water on the back surface to obtain some measure of liquid/vapour transport through the specimen. What data exist implies that transport occurs only by vapour and that most admixtures studied appear to have no effect on the saturated permeability! The liquid/vapour transport in concrete is probably the most pertinent one in practice and yet seems to be the least studied.

31

32 Method IV Kct - (2eEKF)t = θ2This method measures the 'capillary coefficient' K, which in turn is a function of porosity (E), permeability (K) and 'suction1 of the material (F). These variables are also related to the moisture content at any point X from the surface by the equation, Kct - (2eEKF)t = θ2 Where e= density of water t = time interval Kc is not a constant and depends on the initial moisture condition of the concrete. Samples are equilibrated at a chosen relative humidity then weighed and placed in contact with water on one surface only. The amount of absorption is measured by taking weights at various time intervals. Plotting the weight gain against time gives straight lines that may be represented by the equation, Where 'a' represents the amount of water required to initially wet the surface. Slopes of the curves give Kc (g/cm4 X minutes). The capillarity coefficient values with the relative humidity give a Sigmoid pattern and again families of curves have to be compared in order to establish the general and particular effect of an admixture. It is imperative to test against controls.

33 Test methods for waterproofers/ permeability reducersHere we are concerned with ways of measuring the effect of an admixture on altering the flow coefficient Kw or hydraulic permeability. This saturated steady flow coefficient is related to flow rate, etc. by the Darcy formula, Q(cc/sec) = Kw (Δp/ Δx) Kw has units of cm/sec, (Δp/ Δx) is the applied pressure gradient. As with the other flow coefficients Kp and Kc, ad-hoc testing methods have developed due in part to the difficulty in measuring Kw reliably enough to use it as a diagnostic of admixture performance. However, two techniques are described below.

34 Method I As with the other flow coefficients Kp and Kc, ad-hoc testing methods have developed due in part to the difficulty in measuring Kw reliably enough to use it as a diagnostic of admixture performance. However, two techniques are described below. Porous concrete* tablets (127 mm X 51 mm) containing an admixture are cast, cured and at 28 days mounted in an enclosed cell, and water pressure applied (Fig. 5.3(a)). The pressure is increased in stages and maintained for a short period (hours, minutes) at each pressure. The under surface of the tablet is observed for the appearance of moisture droplets. The pressure at which beads of moisture appear is noted and compared against a control sample cast from the same batch of concrete, but containing no admixture. At best this method gives qualitative comparisons and can only relate to short-term behaviour. It is, unfortunately, quite widely used. * A typical porous mix contains 19 mm aggregate, sand and cement in the proportions 3.7 : 2.5 : 1, and has a w/c ratio of about 0.63 to give a 51 mm slump. Air curing in a humidity cabinet is employed.

35 Method II This consists of actually trying to measure Kw and one variant of the apparatus is shown schematically in Fig Concrete specimens are cast as a cylinder or core and then sliced in thicknesses of about cm. The tablets are then set in an annulus of epoxy resin and after the resin has cured are left under water at reduced pressure so as to fully saturate the specimen (Fig. 5.3(b)). After saturation the specimen is fitted into the lower half of a special cell, being sealed with O-rings. The cell is bolted together and filled with water, air being vented from the centre valve and the three-way tap turned so as to fill the calibrated capillary tube (see IS AT method). By isolating the reservoir, movement of liquid along the capillary can be measured which corresponds to flow through the concrete specimen. It is well established that during the early stages of such a test the coefficient of permeability continues to decrease. Many explanations of this effect exist(lO), but from the practical standpoint measurements taken over a seven-day period give results that can be usefully compared. Plotting Kw against time allows performance comparisons to be made.

36