Category Archives: Durability of concrete and cement composites

The purpose of this book is to bring together a series of reviews on topics of current relevance to the durability of concrete and cement-based composites. One might reasonably ask why such a book is now believed to be needed as concrete and related materials have been used with outstanding success in major construction projects for well over two thousand years. Indeed numerous ancient concrete structures, such as the Pantheon in Rome, and masonry structures with mortar joints, such as the Pont du Gard Aqueduct near NõÃmes in southern France (see Fig. 1.1), have survived to the present day in excellent states of preserva- tion. These and other examples of Roman construction works incorporating hydraulic cements made from lime and volcanic earth (or similar vitrified alumino-silicates, known collectively as pozzolanas, after Pozzuoli near Naples where a natural source of such material exists) have shown remarkable dur- ability even where they have been exposed to damp, aggressive environments. An impressive illustration was provided by Davey (1974) whose photograph of part of a Roman breakwater that had been exposed to the sea near Naples for two millennia (reproduced in Fig. 1.2), demonstrates that the mortar joints had endured far more successfully than the now heavily eroded stone blocks which they had been connecting.

Fracture mechanics

Linear elastic fracture mechanics is concerned with predicting conditions that give rise to rapid crack propagation in brittle materials that are considered to be elastic, homogeneous and isotropic at the onset of fracture; it involves the study of stress and displacement at the microscopic level in the region of a crack tip (Mindess and Young, 1981). The field developed from the pioneering work of Griffith (1920) who proposed criteria for the fracture of brittle solids that explained the discrepancy between their low observed bulk strengths and their much higher theoretical strengths (estimated for perfect crystals to be ~ E/10) by postulating the existence of microscopic flaws, which serve as stress concentrators. As illustrated in Figure 3.25, for an elliptical crack with dimensions as indicated, the maximum tensile stress near the crack tip is intensified by a factor that depends on the crack length and end radius according to the relationship:

crack length increases the maximum stress at the tip increases (equation 3.16) while the theoretical stress required to induce energetically favoured crack growth in a brittle material, stress f decreases in accordance with the following expression (Illston et al., 1979):

Similar reasoning has been used to explain the failure of concrete (Kaplan, 1961), which is generally regarded as being brittle with failure occurring at low strains due to linking of existing and newly-formed microcracks.

Strictly speaking, however, concrete exhibits some plasticity (non-linear stress-strain behaviour) prior to failure. It is also evidently heterogeneous as the properties of its constituents are not all identical and it develops a relatively large fracture zone due to microcracking which causes progressive softening of the material prior to complete failure. Various ways of incorporating the above effects into models of the fracture of concrete specimens of different sizes and shapes under different states of applied stress have been proposed and the application of non-linear fracture mechanics is believed to provide a more representative means of describing the fracture behaviour and ultimate capacity of concrete structures. Further consideration of this field is beyond the scope of the present chapter and readers who require more a detailed treatment are referred to specialist works such as the following (ACI SP-118, 1989; ACI Committee 446, 1991; Karihaloo, 1995; Shah et al., 1995).

Degradation of polymer-cement composites Future trends

Interest in polymer-cement composites has increased considerably in recent years with systems finding widespread use in the concrete repair and main- tenance sector and speciality applications such as bridge deck overlays. In these cases, improved properties such as flexural strength, bond to existing material and durability are paramount. Some sectors which currently have a small market should grow and these include areas such as factory-produced pre-cast units, materials that incorporate general wastes or encapsulate hazardous wastes, and materials that, by saving resources, result in a more sustainable environment. High cost tends to preclude their use for more general, high volume applications and the inability to maintain their properties at high temperatures adds a further restriction. Since odour and toxicity of the polymer systems can be a problem, albeit short-term, it would seem prudent to anticipate more stringent restrictions through enhanced environmental legislation.

Relatively small incremental improvements in properties have been made over the years and there is much knowledge and experience available with regard to the various systems and their applications. However, even in established markets, further improvements in properties and reliability are desirable and, for new markets, the improvements would probably have to be substantial to justify cost and environmental considerations. Some questions that come to mind include the following:

· Can we find ways to improve the performance of polymer latexes in both liquid and powder forms (and especially the latter which is more desirable from environmental and quality control points of view)?
· How do the various monomer and polymer systems interact with the cement phases to influence rheology, hydration kinetics and microstructure develop- ment under various curing regimes?
· What determines, especially long-term, the bond between polymer modified cement, silanes and polymer coatings to existing (porous) concrete and aggregate substrates?
· How can we identify the true polymer-cement interface in these complex systems?

The answers to these, and other, questions will only be achieved by research, in particular research that utilises the latest techniques in polymer and cement chemistry and sophisticated instrumental methods of analysis. Procedures in the concrete laboratory should become more scientifically based. Even then this will probably only work if there is good collaboration between the researchers and those supplying the various components of the systems, applying the materials in the field and monitoring their life. A start has been made on standards but more needs to be done so that the properties of materials, as defined in the laboratory, are achieved by recognised application procedures in the field.

 

Characteristics of paints and polymeric surface treatments for concrete

A wide range of products and systems are available and may be classified as pore liners, pore blockers and coatings according to their function, the materials used and method of application. Whilst these systems are generally regarded as beneficial, lack of characterisation of the materials application procedures used makes quantitative comparison between the different products difficult. Similarly, there is little detailed understanding of the mechanisms of how they work or how they may be optimised.

Pore liners

The concrete surface is saturated with organo-silicon compounds such as silanes and siloxanes. In the presence of moisture, these react with the hydrated cement matrix forming a discontinuous film on the surfaces of the pores and capillaries within the concrete, maintaining an open and `breathable’ pore structure. The hydrophobic nature of this chemically modified surface layer reduces moisture ingress and provides good protection against, e.g. chloride ion penetration (Basheer, 1998). Depth of penetration achieved in a given time depends on such factors as pore size distribution of the mortar matrix, water content and the composition and reactivity of the compound being impregnated (Sosoro, 1998; Basheer, 1998). Whilst formulations of silanes/siloxanes are popular, there is some concern regarding their toxicity and this has instigated investigations into more environmentally friendly alternatives (Chamberlain, 2005). One such solution is water-based and after application is said to precipitate hydrophobic crystals that adhere to the pore walls of the concrete. Although details of the material used are vague the treatment is said to conform to Highways Agency Standard (BD 43/03, 2003).

Pore blockers

These materials essentially harden the surface of the concrete and reduce its porosity by partially or totally filling the pores and capillaries of the cement matrix. Silicates and silicofluorides, which react with any free lime, are available for this purpose although the polymers used for polymer impregnation described in Section 10.4 may be used. Compared to pore liners, there is little penetration of the concrete surface, (Basheer, 1997).

Coatings

The aim here is to produce a continuous protective layer on the surface with a typical thickness of 0.1±5.0mm depending on the application. Coatings function by sealing the surface of the concrete against water absorption, whilst allowing transmission of water vapour (Dietricht, 1998) and ingress of carbon dioxide and chloride ions. They should also be effective in crack bridging. A range of systems are available with epoxy and polyurethane performing better than acrylic, chlorinated rubber and polymer emulsion coatings (Almussallam, 2003) although noticeable variations in performance of the same type but from differ- ent sources was found. Some results (Ibrahim, 1999) suggest that a combination of silane treatment plus an acrylic top coat was very effective in reducing carbonation and chloride ion ingress. Presumably in this case the silane acts a coupling agent between the polymer and concrete.

Guidance on the use of these systems is given in two Standards (BS EN 1504-1: 1998 and BS EN 1062-1: 1997) with the approach being described by Hurley (2004).

Degradation of polymer-cement composites Interfacial characteristics

Although a large amount of data have been obtained for a range of polymer- cement composite properties, relatively little is known, apart from that discussed  in Section 10.2.4, about the nature of polymer-cement interfaces and interactions that may take place between the two components. In the field of composites generally, it is widely accepted that interfaces between different components are important in determining properties such as strength and durability. It is reasonable to assume that this will also be true in the case of polymer-cement composites. Studying the interface between the cement and polymer phases in a com- posite involves a number of practical difficulties. As the cement phase is porous, it is quite probable that polymer will penetrate surface pores resulting in a complex and diffuse interfacial layer. The actual interfacial region itself is hard to reveal and if interaction products have been formed, amounts may be very small and difficult to extract and analyse separately from the cement or polymer phases. Different specimen geometries may be mechanically loaded with the aim of achieving at least some interfacial failure thus permitting subsequent surface examination of the interfacial region. This is, however, difficult to achieve in practice even with carefully notched samples with the locus of failure often passing through bulk material. An alternative method is the selective dissolution of either the polymer or cement phase, in theory leaving the interfacial region available for surface analysis. In addition, the amounts of interaction products may be magnified by increasing the surface area, e.g. using a powdered cement rather than trying to cast against a flat surface. A number of systems have been investigated (Short, 1999; Shaw, 1989) with these points in mind, concentrating on the case when the monomer, methyl methacrylate (MMA), comes into contact with hydrated cements and is subsequently cured to give poly(methyl methacrylate) (PMMA). It was found that the monomeric MMA reacted with the cement to form primarily calcium methacrylate. This reaction proceeds in two stages:

Since the calcium methacrylate formed is highly water soluble, a reaction such as this would be generally detrimental to the properties of a polymer± cement composite. Water is a limiting reagent in these reactions and the amount of methacrylate formed is proportional to the initial amount of water present in the cement. In practice, if the highly water-soluble methacrylate were leached  from the polymer±cement interface over time then the properties of the composite would be modified. This confirms the need for thorough drying of the cement or aggregate prior to contact with the monomer if adhesion is to be maximised.

Polymer impregnated concrete

A polymer impregnated concrete is produced via a process whereby an existing hardened concrete (structure or factory component) is dried at around 150ëC and then impregnated with a low viscosity monomer, usually methyl methacrylate, which is subsequently cured in situ. Impregnation is often achieved by soaking at atmospheric pressure although the process may be facilitated by evacuation and/or higher pressures. Polymerisation is accomplished by a thermal catalytic or promoted catalytic process or is radiation induced. The resulting polymer fills, or lines the surface, of the pores of the concrete modifying the properties of the hardened mortar. Extensive reviews of early work (Swamy, 1979; Shaw, 1989) showed that increases in modulus of elasticity, strength (compressive, flexural and tensile) and improvements in durability could be effected, although the extent of these improvements does depend on the depth of impregnation achieved and porosity of the original concrete. The process shows the most marked improvements when starting with high porosity, low compressive strength concretes. Since these reviews were written, there has been only a little additional investigative work reported (Ohama, 1997; Fowler, 1999). More recent work (Chen, 2005) has compared soaking time and polymerisation tem- perature on the mechanical properties of PIC in a systematic way and confirmed the possible improvements. Decreases in surface absorption compared with normal concrete are attributed to significant decreases in total porosity with maximum pore diameters of <50 nm. Despite the much improved properties and early application of PIC for remedial treatment of road bridge decks in the US, their high processing costs has generally precluded further use apart from limited precast operations in Japan, (Ohama, 1997).

However, one developing area where polymer impregnation may prove beneficial is the wear resistance of concrete used in floor construction. Treat- ments have been found to significantly increase the abrasion resistance of all types of concrete (Sadegzadeh, 1988), although improvements are greatest with low compressive strength concrete (Sebok, 2004).

Reactive polymer matrix Thermal properties

In general the coefficients of thermal expansion for unfilled (lower modulus) and filled (higher modulus) products are ~40-100 and 20-40  10^-6 C^-1 respec- tively. This compares with ~10 x 10^-6 C’^-1 for typical substrate materials and can lead to the development of significant thermal stresses. Thus for high modulus systems which undergo little stress relaxation, temperature changes ~30ëC can lead to stresses which, over time, can lead to shear failure in a concrete surface, loss of adhesive bonding from a steel substrate or cracking of the resin.
Closer matching can be obtained by careful attention to filler type and loading, and obtaining the best balance between thermal expansion coefficient, modulus and stress relaxation.
The fire performance of these resin-based materials is inferior to steel or concrete and must be taken into account when considering potential applica- tions. Whilst this factor would preclude their use in a few situations, there are many where it is of little consequence.

Reactive polymer matrix composites Creep

These materials are visco-elastic in nature and thus, over time, may undergo creep and stress relaxation even at ambient temperatures. For many applications, service stresses are low compared to ultimate values and the effects are limited. However, for applications involving adhesive bonding and other load bearing applications, the consequences may have to be considered. Thus, it may be necessary to determine the limiting stress level below which creep rupture does not occur for applications such as anchoring. In complex loading systems such as shrinkage or differential thermal expansion and contraction, stress relaxation may alleviate stress concentrations. The extent of these effects is governed by factors such as degree of cross-linking in the resin and the type/level of filler, see for example (Mays, 1992).

Reactive polymer matrix composites Water absorption

Reactive polymer matrix composite materials generally have a high resistance to water transmission, although the extent depends very much on their formulation. Over the long-term, slow diffusion may lead to reversible equilibrium moisture contents of up to 10%. Eventually this can lead to the creation of pathways for ingress of deleterious species such as chloride ions, reduction in modulus and strength by acting as a plasticiser and degradation of the resin by hydrolysis. These effects are increased by air entrainment, due to poor mixing or compac- tion, and as temperatures are increased. Moisture ingress via the polymer, a porous substrate or exposed edges can lead to degradation of the adhesive bond.

The mechanism is not fully understood but probably involves plasticising and swelling of the polymer in the interfacial region and alkaline hydrolysis of the resin.

Reactive polymer matrix composites Application

Since there is a very wide range of systems available, it is very important that the right one is chosen for the specific application in mind. Assuming that the correct choice is made, it is then very important that the correct procedures are used on site and in the case of pre-cast factory products.

Correct proportioning and thorough mixing of resin and activator or catalyst is required in order to achieve the correct curing rate and this must also be consistent throughout the product. Application should also be within the speci- fied temperature range and sufficient time allowed for the desired properties to be achieved. For example, incorrect mixing can result in some areas being hard whilst other parts remain soft or rubbery.

On-site use invariably requires a good bond to an existing concrete or steel substrate and this can only be achieved if the surfaces are clean and any loose material or corrosion products are removed. Good interfacial contact is required which can usually be achieved by the use of primers. In addition, whilst some systems are formulated for application to damp surfaces, the majority require that the surface moisture content be reduced to <5% during application and cure. Capillary movement of moisture from deeper within the substrate may occur later and lead to reduction of bond or blistering in the case of floorings and coatings. Little is known about the factors driving this process but they probably involve the pore size distribution of concrete, presence of interfacial soluble compounds and the source of moisture. Osmotic blistering may occur on steel and concrete substrates through water ingress as a result of salts on the steel surface (poor cleaning) or, in the case of concrete, the use of incorrect primers containing water-soluble compounds, (Dively, 1994; FeRFA, 1990). In addition to the above, long-term adhesion depends on factors such as level of applied stress, fatigue, exposure to moisture, and temperature fluctuations.

Reactive polymer matrix composites Introduction

These materials are quite different from polymer-modified mortars in that they do not contain any hydraulic cement. They consist of a synthetic resin binder incorporating a hardener and filler such as well-graded powder, sand or coarse aggregate. These are mixed on site, placed and then allowed to cure giving a material with generally good mechanical properties (in terms of compressive, tensile, flexural strengths), adhesive bond strength, and chemical and abrasion resistance. Applications of these materials include repair mortars and crack injection formulations, adhesives, grouts for mechanical fixings, floor finishes such as sealers and toppings, and pre-cast drainage channels and pipes. Several recent and detailed reviews are available (CIRIA, 2000).

The most widely used resins include those of the following generic types: epoxies, unsaturated polyesters, methacrylates, polyurethanes, vinyl esters and furans. Their chemical nature allows the formulation of a wide variety of products ranging from low viscosity solvent systems for void impregnation and crack injection to heavily filled products much like conventional mortars.

Thermosetting resins are used so that, when combined with curing agents or catalysts at ambient temperatures, they undergo further polymerisation (cross- linking) to give a three dimensional structure. The correct degree of cross- linking is very important in that it determines the glass transition temperature of the polymer and other long-term properties.

The amount of curing agent or catalyst governs the rate of reaction at a given temperature so that, if the material were applied on site at low ambient tem- peratures of say ~10ëC, more is required, whilst at higher ambient temperatures of say ~35ëC the converse is true. In addition, the reactions are influenced by the exothermic nature of polymerisation, the thickness of the coating and variations in ambient temperature. Most reactions do not proceed to completion but, if reactive polymer matrix composites are formulated appropriately, a sufficient  degree of cross-linking is attained at ambient temperatures to give the required service properties.