Concrete Is A Dynamic Material

Published Date: 02 Nov 2017

2.1 Introduction

Concrete is a dynamic material, one whose microstructure changes with time and one whose volume will change from the early stages of hydration to the later stages of service life. Concrete structures do not frequently fail due to lack of strength, rather due to inadequate durability or due to improper maintenance techniques. The most common cause of premature deterioration is attributed to the development of cracks (Mehta, 1992; Hobbs, 1999).

Cracking can occur in concrete pavements and structures for numerous reasons that can primarily be grouped into either mechanical loading or environmental effects. It should also be noted that for most practical structures, reinforcement is used to bridge and hold cracks together when they develop, thereby assuring load transfer while adding ductility to a relatively brittle material. Therefore not all cracking causes concern. Reinforced concrete elements are frequently designed on the assumption that cracking should take place under standard loading conditions (Nilsson and Winter, 1985; Nawy, 2000). For example continuously reinforced concrete pavements (CRCP) are designed with longitudinal steel in an amount adequate to hold shrinkage cracks tight, while joints exist only at locations of construction transitions and on-grade structures. In this pavement type wherein shrinkage cracks develop over time and stabilize over the first 3 to 4 years, cracking in the transverse direction in specific patterns is not detrimental to the structure as long as the cracks remain tight and retain good load transfer. Therefore, cause of cracking should be carefully identified to determine which cracks are common and acceptable and which cracks merit repair or further investigation. Mechanical loads induce strains that can exceed the strain capacity (or strength capacity) of concrete, thereby causing cracking. Concrete may be particularly susceptible to cracking that occurs at early-ages when concrete has a low tensile capacity (Kasai, 1972). If the loads are applied repeatedly or over a long period of time, fatigue and creep can affect the strain (or strength) development that can lead to failure (Bazant and Celodin, 1991) or reduce stresses (Shah et al., 1998). Although numerous factors influence whether concrete would be expected to crack due to environmental effects, it can be simply stated that cracking will occur if the stress that develops in response to internal expansion or the restraint of a volumetric contraction that results in stress development exceeds the strength (or fracture resistance) of the material. Internal expansion is primarily caused by chemical attack or freezing of the pore water while volumetric contraction is typically attributed to moisture changes, chemical reactions, and thermal changes.

2.2 Types of Cracks

Cracking can be classified based on when they appear on concrete, before hardening and after hardening. Cracks that occur before hardening are primarily due construction movement, premature freezing, settlement and excessive evaporation of water that is plastic cracks. Plastic cracking can be eliminated through close attention to the mixture design. Cracks that occur after hardening may be due to volume instability, mechanical loading or chemical reactions of the incompatible materials e.g. alkali-aggregate reactions. Figure 2.1 distinguishes the types of cracks based on before and after hardening (TRC, 2006).

Creep

Fatigue

Sub-grade Movement

Formwork Movement

AAR/ASR/DEF

Steel Corrosion

Freeze-Thaw cycling

Design/Sub-grade

Design Load/Overload

Thermal Change

Drying Shrinkage

Scaling, Crazing

Premature Freezing

Autogenous Shrinkage

Plastic Settlement

Plastic Shrinkage

Types of Cracks

Volume Instability

Physio-Chemical

Structural Design

Frost

Damage

Plastic

Construction Movement

After Hardening

Before Hardening

Figure 2.1 Common types of cracking in concrete structures

2.3 Plastic Shrinkage

Plastic shrinkage appear in the surface of fresh concrete at early age and while it is still plastic. These cracks appear mostly on horizontal surfaces, they are usually parallel to each other. Plastic shrinkage depends on two primary factors; the rate at which surface water forms (bleeding) and the evaporation rate of the surface water (Wang et al., 2001). When the evaporation rate from the top surface of the concrete exceeds the bleed rate at which water rises from the concrete, the top surface dries out. At this point, the receding water below the concrete surface forms menisci between the fine particles of cement and aggregate causing a tensile force to develop in the surface layers. If the concrete surface has started to set and has developed sufficient tensile strength to resist the tensile forces, cracks do not form. If the surface dries vary rapidly, the concrete may still be plastic and cracks do not develop at that time, plastic cracks will surely form as soon as concrete stiffens a little more.

The rate of evaporation from the surface depends on environmental factors such as temperature, relative humidity and wind speed. It is not just the hot weather phenomenon, as the combination of these factors may provide the worst condition in cool weather with low humidity and wind. The rate of evaporation can be determined from the relative humidity, air temperature, concrete temperature and wind velocity using the nomograph in figure 2.2 or the equation developed by Uno. Cracking is most likely to occur when the environmental conditions give an evaporation rate in excess of 1 kg/m2/h. It is recommended that precautions should be taken when the anticipated evaporation rate is likely to exceed 0.5 kg/m2/h. Uno gives the following equation to calculate evaporation rate.

E= 5([TC + 18]2.5 – r[Ta + 18]2.5)(V + 4) X 10-6 Eq. 1

Where;

E = evaporation (kg/m2/h)

r = relative humidity/100

V = Wind velocity (km/h)

Ta = air temperature (oC)

Tc = Concrete (water surface) temperature (oC)

NOTE: Temperature, humidity and wind velocity need to be measured on site to give a realistic picture of the evaporation conditions.

Both the nomograph and the equation are based on evaporation from a water surface and does not hold true after bleed water have disappeared from the surface.

2.3.1 Causes of Plastic Shrinkage

Plastic shrinkage occurs when the rate of evaporation of moisture from the surface exceeds the rate at which moisture is supplied to it. The concrete surface dries out and shrinks at a time which it has little strength and it cracks. Water is lost from the concrete mass in two ways which are drying from the top and drying from the base. Drying from the top occurs when the moisture rises to the top surface of a concrete element during placement. Bleed water dries out due to evaporation and when the rate of evaporation exceeds the rate of bleeding, the surface dries and ends to crack. Drying from the base occurs when water in a concrete slab is absorbed into the ground or sub-grade below. This could increase the settlement of the concrete significantly and cracking can occur.

USE OF CHART

1. From air temperature move UP to relative humidity.

2. Move RIGHT to concrete temperature.

3. Move DOWN to wind speed.

4. Move LEFT to read the rate of evaporation.

Figure 2.2: Effect of ambient conditions on rate of evaporation (ACI 2005)

2.4 Plastic Settlement

Plastic settlement cracks are so-called because they form while the concrete is still plastic. Plastic settlement crack occurs in freshly mixed concrete as the concrete settles over time and encounters some restraint. The heavier particles sinks due to gravity until the concrete sets. The settling concrete is restrained and cracks form at the surface. They may become visible very early, that is while finishing is proceeding, but often not noticed until some hours after placement. They are distinguished from plastic shrinkage cracks by their distinct pattern which typically mirrors the pattern of the restraining elements such as the reinforcement.

The cracks occur while the concrete is plastic and frequently while bleed water is still rising and covers the surface. They tend to roughly follow the restraining element, e.g. reinforcing bars, or changes in the concrete section. They can be quite wide at the surface, tend to extend only to the reinforcement or other restraining element and taper in width to that location. The cracks can be further exacerbated by in adequate compaction and the presence of void under reinforcing bars. In exposed situations this may increase the risk of corrosion of the reinforcement and pose a threat to durability. Cracks may develop further, due to subsequent drying shrinkage, leading to possible cracking through the full depth of the concrete member (G.C.C., 2002).

2.4.1 The Mechanism of Plastic Settlement

After the placement of concrete, the solids in the concrete settles down and the mix water rises up to the surface. When solids in the concrete settles freely without hindrance there will be a reduction in depth and volume of the cast concrete but cracking will not occur. However, any restraint to the movement e.g. reinforcement, can result in plastic settlement crack (John Newman, 2003). Where the solids continue to settle in comparison to those which are prevented from further downward movement, the concrete back will break and a tear appears in the surface as it is forced into tension. Cracks may develop at regular spacing reflecting the reinforcement layout. They often occur in conjunction with voids under the reinforcing bars as shown in figure 2.3.

Figure 2.3; Formation of plastic settlement crack (John Newman, 2003).

These crescent-shaped voids may initially be filled with bleed water. The region of bond between the bar and concrete is thus reduced. When the restraint occurs near to the surface, formation of cracks is more likely to occur. A settlement crack is unlikely to occur if the depth of cover to the reinforcement is greater than one third of the section depth (Turton, 1981). The rate of evaporation (wind speed) and mix proportions (tendency to bleed) would be expected to affect the severity of the cracking. The number of cracks is influenced by the occurrence of the restraint. However, the reinforcement diameter and concrete workability have little influence.

2.5 Chemical Shrinkage

Chemical shrinkage can be defined as "the phenomenon in which the absolute volume of hydration products is less than the total volume of unhydrated cement and water before hydration." (Tazawa et. al., 1999). This type of shrinkage is a result of the reactions resulting between cement and water, which leads to a volume reduction at early age. However when the concrete begins to stiff, chemical shrinkage tends to create pores within the mix structure. Chemical shrinkage is based merely on the volume of initial and final products. It can be calculated based on molecular weights. The general equation is given below.

Vcs = Eq. 2

Where;

Vcs = Chemical shrinkage

Vc = Volume of hydrated cement,

Vw = Volume of reacted water,

Vhy = Volume of hydrated products,

Vci = Volume of cement before mixing,

Vwi = Volume of water before mixing.

Table 2.1 shows the result of shrinkage on cement phases by Pauliuni (1996). The study shows that dicalcium silicate (C2S) shrinks the least. This is likely due to surface reactions of the C2S and its polymorphic behavior or the presence of retarding agents. Tricalcium alumimate (C3A) shrinks the most because it reacts extremely fast in the first hours. While tricalcium aluminate (C3S) and tetracalcium alumino ferrite (C4AF) are found to be gradual over many days.

Table 2.1; Cement composition influence on Chemical Shrinkage (Paulini, 1996).

Shrinkage (cm3/g)

Phase

1 day

3 days

7 days

14 days

28days

C3S

0.0188

0.0300

0.0336

0.0406

0.0481

C2S

0.0110

0.0126

0.0106

0.0140

0.0202

C3A

0.0632

0.0759

0.1133

0.01201

0.1091

C4AF

0.019

0.0202

0.0415

0.0352

0.0247

Chemical shrinkage begins immediately after mixing water and cement. The magnitude of chemical shrinkage can be determined using the molecular weight and densities of the compounds as they change from basic reaction of the products.

Chemical shrinkage can be measured directly using two methods: dilatometry and weighing reduced buoyance. The dilatometry test procedure follows Le chateliers principle where an Erlenmeyer flask containing a diluted cement pastesample is connected to a pipette from which the droping water level is monitored. The cement paste must be diluted to provide a high enough permeability for water to penetrate and saturate the entire capillary pores (Gange et al. 1999). The weighting reduced buoyancy method is based on Archimedes principle where the volume reduction of the submerged sample is monitored through an apparent weight increase. The mass of the cement paste will lower as the cement hydration proceeds.

2.6 Autogenous Shrinkage

Autogenous shrinkage can be defined as the microscopic volume change occurring with no moisture transferred to the exterior surrounding environment and it is a result of chemical shrinkage affiliated with hydration of cement particles (J.C.I., 1999). Autogenous shrinkage occurs even when the concrete is completely submersed in water, thus having 100% humidity on the surface. It also occurs even when the surface is made completely air and water proof with some curing agent. Thus its mechanism is not related to surface tension of water at the surface, but rather to the surface tension in pores, a reduction in relative humidity as the pore water is chemically consumed, and the actual volume change from the reactants to the products (Xi et al., 2003).

Figure 2.4 shows a graphical representation of a sealed concrete composition change due to cement hydration reaction. It also indicates that autogenous shrinkage is a portion of chemical shrinkage. Chemical shrinkage is an internal volume reduction while autogenous shrinkage is an external volume change. It is therefore possible to measure autogenous shrinkage as a linear change on a concrete slab.

Figure 2.4; Reactions causing autogenous and chemical shrinkage (J.C.I., 1999).

C = unhydrated cement, W = unhydrated water, Hy = hydration products, and

V = voids generated by hydration.

The first source of chemical shrinkage is from volume reduction of the reaction products. This is dominant at a very early age, when the concrete is still liquid. At this age the chemical and autogenous shrinkage does not result in stress, as the concrete is unrestrained and simply settles.

The factors influencing the magnitude of autogenous shrinkage are often disputed. It is agreed that autogenous shrinkage cannot be prevented by casting, placing or curing methods, but must be addressed when proportioning the concrete mixture. It appears that the internal components or ingredients have the most significant influence, as will be discussed in upcoming paragraphs. This volume change has been given a variety of labels, some of which include: autogenous deformation, chemical shrinkage (both total and external), volume contraction, Le Chatelier shrinkage, bulk shrinkage, indigenous shrinkage, selfdesiccation shrinkage, and autogenous volume change. (Justnes et al. 1996).

According to Lyman (1934) autogenous shrinkage was described in the 1930`s as a factor contributing to total shrinkage which was difficult to assess. In these earlier days, autogenous shrinkage was noted to occur only at very low water-to-cement ratios that were far beyond the practical range of these concretes. But with the development and frequent use of modern admixtures, such as super plasticizers and silica fumes, it is much more realistic to proportion concrete susceptible to autogenous shrinkage. Today we often have greater structural demands for high strength and high performance concretes. This leads engineers and designers to specify concrete with lower w/c ratios, much beyond the limitations of the 1930’s. Even though many strength and durability aspects are now improved with these specifications, the risk of autogenous shrinkage is greater.

Autogenous shrinkage takes place over three different stages within the first day after concrete mixing: liquid, skeleton formation, and hardening. After the hardening stage the concrete shrinkage can be measured using more standard long term measuring practices. During the first day it is necessary to clearly define the mechanisms that are generating the autogenous shrinkage in these three stages.

2.6.1 Liquid Stage

Immediately after mixing water and cement, a chemical shrinkage change will occur due to the reduction in volume of the reaction products. In this early phase while the concrete is still liquid the autogenous shrinkage is equivalent to chemical shrinkage.

2.6.2 Skeleton Formation Stage

The point when autogenous shrinkage changes from being a function of chemical shrinkage versus self-desiccation is a function of the degree of cement hydration. Self-desiccation is defined as the localized drying of the concrete’s internal pores. At the point when a skeleton is formed due to the stiffening of the paste the concrete can resist some of the chemical shrinkage stresses. Soon after this initial skeleton formation the concrete will set. During this stage the capillary pressure will start to develop and may cause shrinkage. This pressure mechanism works as the water, or meniscus, is moving between the pores. As the water is lost from subsequently smaller pores, the water meniscus will continue to be pulled into the capillary pores and will generate more stress on the capillary pore walls.

Figure 2.5 shows the changes occurring during skeleton formation according to Acker (1988). When the concrete is still liquid immediately after mixing the autogenous shrinkage is proportional to the degree of hydration (section AB in the figure). This means the autogenous shrinkage is due only to chemical changes. Once a skeleton has formed the chemical shrinkage becomes more and more restrained (Section BC). Beyond point C the material is rigid and autogenous shrinkage is comprised of less and less chemical shrinkage. The further volume reductions are only due to self-desiccation.

Figure 2.5; Evolution of autogenous shrinkage as a function of degree of hydration. (Acker 1988)

These changes in the driving force of autogenous shrinkage over time have also been well documented by Hammer (1999) and Justnes et al. (1996). They simultaneously conducted a volumetric tests of chemical shrinkage (bottle test) and autogenous shrinkage (thin rubber membrane test) on cement paste, as shown in Figure 2.6. Here it can clearly be seen that once a skeleton starts to form at approximately 5 hours the autogenous shrinkage diverges from the chemical shrinkage. But during the very early hours the autogenous shrinkage can be fully attributed to the chemical changes due to cement hydration.

Figure 2.6; Chemical shrinkage by bottle test and volumetric autogenous shrinkage by thin rubber membrane test. Cement paste with water-to-binder ratio of 0.40 and 5% silica fume. (Hammer 1999)

2.6.3 Hardened Stage

Once concrete has hardened with age greater than one day, the autogenous shrinkage may no longer be a result of only chemical shrinkage. During the later ages the autogenous shrinkage can also result from self-desiccation since a rigid skeleton is formed to resist chemical shrinkage. Self-desiccation is the localized drying resulting from a decreasing relative humidity. The lower humidity is due to the cement requiring extra water for hydration.

In a high strength concrete with a low w/c ratio, the finer porosity causes the water meniscus to have a greater radius of curvature. These menisci cause a large compressive stress on the pore walls, thus having a greater autogenous shrinkage as the paste is pulled inwards.

Self-desiccation occurs over a longer time period than chemical shrinkage and does not begin immediately after casting. It is only a risk when there is not enough localized water in the paste for the cement to hydrate; thus the water is drawn out of the capillary pore spaces between solid particles. This would typically begin after many hours or days in high strength (low w/c ratio) concretes.

Research by Powers & Brownyard (1948) showed that autogenous shrinkage due to self-desiccation occurs when the w/c ratio is below 0.42, since all mixing water is consumed at this ratio. This w/c limit can vary from 0.36 to 0.48 (Taylor 1997), depending on the cement type, but the most common ratio cited in literature is 0.42. (Neville 1978, Mindess & Young 1981) The lower values (i.e. 0.38) assume there is an unrestricted supply of water available during curing. When the w/c ratio is much lower than 0.42 and can no longer gain curing water, the cement will seek extra water from the internal pores and thus lower the relative humidity.

The use of mineral admixtures, such as silica fume, may also refine the pore structure towards a finer microstructure. If there is a finer pore structure the water consumption will be increased and thus the autogenous shrinkage due to self-desiccation will be increased. On the other hand, research by Kinuthia et al. (2000) has shown that the mineral admixture Metakaolin may reduce the autogenous shrinkage caused by self-desiccation by improving the particle size distribution and causing expansion that compensates shrinkage.

Work by Igarashi et al. (1999) has shown that adding aggregate to a mixture will decrease the amount of autogenous shrinkage. This seems reasonable, since the aggregates are rigid and cement paste is the portion of the mixture responsible for the shrinkage. His tests showed concrete shrinkage 30 to 45% less than the shrinkage corresponding to pastes at the age of a few days.

The draft European Standard prEN 1992-1 [Eurocode 2001] is the first to include a method for predicting long term autogenous shrinkage strain, as given by Equation 3. The strain is based on the cement type and compressive strength of the concrete.

Ɛcs = βcc(t)Ɛcs∞ Eq. 3

Where

Ɛcs = Autogenous shrinkage

βcc(t) = 1 – exp (-0.2t0.5)

Ɛcs∞ = 2.5(fc – 10) x 10-3

fc’ = Compressive strength of concrete at 28 days.

t = Age of concrete in days

In a high strength concrete element there will be no variation in the autogenous shrinkage magnitude over the cross-section or depth. Since the porosity is low and no excess curing water can infiltrate high strength concrete, the internal water will be evenly distributed after concrete hardening and reaching temperature equilibrium. The loss of internal water will be localized on a microscopic scale in the capillary pores and will not register as a change in autogenous shrinkage magnitudes. This is much different than drying shrinkage, where the amount of shrinkage is dependent on the moisture gradient across the element generated by uneven evaporation.

2.7 Drying Shrinkage

Drying shrinkage is the most significant type of shrinkage in most concrete mixes, and has been called the most deleterious property of Portland cement composites (Zhang and Li, 2001). The mechanisms are similar to that of plastic shrinkage but this occurs after hardening. When concrete is exposed to its service environment, it tends to reach equilibrium with that environment. If the environment is a dry atmosphere the exposed surface of the concrete loses water by evaporation. The rate of evaporation depends on the relative humidity, temperature, water-cement ratio and the area of the exposed surface of the concrete. The first water to be lost is that held in the large capillary pores of the hardened concrete. The loss of this water does not cause significant volume change. However, as drying continues, loss of water from small capillary pores and later from gel pores takes place. With the reduction in the vapour pressure in the capillary pores, tensile stress in the residual water increases. Tensile stresses in the capillary water are balanced by compressive stresses in the surrounding concrete and as a result the concrete shrinks. Evaporation of gel water changes the surface energy of the solid phase and causes further shrinkage (C.C.A.A., 2002). Figure 2.7 shows various components of deformation of in a concrete member.

Figure 2.7; Various components of deformation

Figure 2.8; Concrete subjected to drying and wetting

If the environment is dry, the exposed surface of the concrete loses water by evaporation. But if the environment is wet, the movement of moisture will be from the environment to the concrete this will result in volume increase or swelling. Figure 2.8 shows a schematic description of volume change in the concrete due to alternate cycles of drying and wetting.

2.7.1 Factors Affecting Drying Shrinkage

Drying shrinkage is influenced by internal factors related to the concrete and its constituents and external factors that affect drying.

2.7.1.1 Internal Factors

The internal factors affecting drying shrinkage of concrete are related to cement, aggregates, admixtures, water-cement ration, water content, placing compaction and curing.

Cement

The effect of cement composition on drying shrinkage is not completely determined. The C3A and alkali content on shrinkage have been observed to have a dominant effect, and it is influenced by gypsum content of the cement, i.e. shrinkage of cement of the same C3A content differs for different gypsum contents (Pickett G., 1974).

Aggregates

Aggregates have a restraining effect on shrinkage. According to Welch G. B. (1978), if the aggregate shrinks less than the paste then the aggregate restrains shrinkage and the shrinkage of the concrete will decrease with increase of the aggregate volume fraction. Aggregate properties such as grading, maximum size, shape and texture affect drying shrinkage indirectly. These properties can lead to a change in the water or paste content and their effect on drying shrinkage can only be measurable in terms of the changes they cause to the concrete mix.

Water Content

The increase in water content also raises the drying shrinkage of concrete. Concrete of high water content has a lower strength and lower modulus of elasticity and has a significant tendency to shrinkage. Figure 2.9 shows the effect of water-cement ratio on drying shrinkage.

Figure 2.9; Effect of water-cement ratio on shrinkage (Soroka I., 1979).

2.7.1.2 External Factors

The external factors affecting loss of moisture from concrete are ambient conditions and size and the shape of the concrete member.

Ambient Condition

Air temperature, relative humidity, and wind velocity will affect the loss of moisture from the concrete surface. Drying shrinkage is high when there is a rise in ambient temperature, relative humidity decreases, air movement around the concrete increases or increase in the length of time for which concrete is subjected to drying conditions (Blakey F.A., 1963). Figure 2.10 shows the effect of relative humidity on drying shrinkage of concrete.

Figure 2.10; Effect of relative humidity on drying shrinkage (Blakey F.A., 1963)

2.8 Thermal Effects

Thermal effects are as important to the cracking problem as shrinkage is, but often overlooked since they are outside the control of the engineer. Nevertheless the strain applied by temperature alone can easily be enough to cause cracking (Krauss and Rogalla, 1996; Aktan et al., 2003). As with any material, concrete expand when heated and contracts when cooled. If these deformations are restrained, an increase in temperature produces compressive stresses while a decrease in temperature produces tensile stresses. Figure 2.11 shows an illustration of the development of concrete temperatures and thermal stresses over time for a freshly placed concrete. In regards to stress development, the final-set temperature is the temperature at which the concrete begins to resist stresses induced by restraint of external volume changes.

In figure 2.11, it can be noted that due to hydration, the concrete temperature increases beyond the setting temperature, line (A). As the expansion of the concrete caused by the temperature rise is restrained, the concrete will be in compression when the peak temperature is reached, line (B). The phenomenon of gradual increase in stress over time is called relaxation. When the peak temperature is reached, its strength is low and high amounts of early-age relaxation may occur when the concrete is subjected to high compression loads. When the concrete temperature decreases, compressive stress is gradually relieved until the stress condition changes from compression to tension, line (C). The temperature at which this transient stress-free condition occurs is denoted the "zero-stress temperature." Due to the effects of relaxation, the zero-stress temperature may be significantly higher than the final-set temperature. If tensile stresses caused by a further temperature decrease exceed the tensile strength of the concrete, cracking will occur, line (D). Because the thermal stress is proportional to the difference between the zero-stress temperature and the cracking temperature, thermal cracking can be minimized by decreasing the zero-stress temperature. This in turn can be accomplished by minimizing the final-set temperature, minimizing the peak temperature achieved during the high-relaxation phase, or delaying the attainment of the peak temperature (Byard et at., 2010).

Figure 2.11; Development of thermal stresses (Byard et at., 2010).

The magnitude of thermal stress development depends on the amount of temperature change as well as several other factors, namely the coefficient of thermal expansion, the creep-adjusted modulus of elasticity, and the restraint conditions (Bamforth and Price 1995). These factors are represented mathematically in Equation 4.

THERMAL = T x α(t) x Ecr (t) x Kr Eq. 4

Where,

THERMAL = change in concrete stress,

T = change in concrete temperature

α(t) = Coefficient of thermal expansion (strain/oc)

Ecr (t) = Creep adjusted module of elasticity (Pa)

Kr = restraint factor

The coefficient of thermal expansion and the creep adjusted module of elasticity are time dependent. In order to accurately compute thermal stresses, comprehension of the complex interactions of the variables is of utmost importance.

2.9 Factor Affecting Early Age Shrinkage

The parameters that affect early-age shrinkage of concrete differ according to the driving forces behind each shrinkage mechanism. Drying shrinkage is mainly attributed to environmental conditions. Thus, problematic factors are those that affect the rate of evaporation, i.e. relative humidity, air velocity, air and concrete temperature (Esping, 2007). Additionally, the materials used in concrete have a parallel effect through controlling the quantity and duration of bleeding and setting time (Holt and Leivo, 2004). On the other hand, autogenous shrinkage is influenced by the type and properties of the binder, mixture proportions, and admixtures that can refine the pore structure. Although the early-age autogenous shrinkage is fully attributed to chemical shrinkage, it may behave differently than chemical shrinkage with respect to some factors especially those that affect the setting time and formation of a restraining structure (Esping, 2007).

Generally, the early-age shrinkage is affected by the following parameters:

i) The binder type, content and rate of hydration: A higher rate of hydration results in higher autogenous and drying shrinkage due to the decreased volume of hydration products relative to their constituents and the higher water consumption, which in turn reduces bleeding and increases the concrete temperature (Bentur, 2003, Holt, 2005, Tazawa, 1999, Esping, 2007, Topçu and Elgün, 2004).

b) The aggregate content: Aggregates reduce shrinkage, and act as an internal restraint. It also reduces the volume of cement paste, leading to lower chemical shrinkage (Holt, 2001, Esping, 2007). Furthermore, light-weight aggregates with high absorption were found to reduce autogenous shrinkage as it acts as an internal curing source (Nassif et al., 2003, Mihashi and Leite, 2004, Tazawa, 1999, Zhutovsky et al., 2002, Duran- Herrera et al., 2007).

c) The water content: it has a major role during early-age shrinkage through controlling the amount of free water, and the development of the microstructure and pore system, which consequently affects the capillary tension and meniscus development (autogenous shrinkage).

d) Admixtures: a shrinkage reducing admixture (SRA) can decrease the surface tension of the capillary pore solution resulting in a reduction of the capillary tension. It was found to reduce both autogenous and drying shrinkage strains (Bentur, 2003, D’Ambrosia, 2002, Holt, 2005, Esping, 2007, Lura et al., 2007, Bentza et al., 2001).

Reduction of about 60% was observed in the early-age unrestrained shrinkage due to an SRA dosage of 1- 2% by mass of the cement within the first week after casting concrete

(D’Ambrosia, 2002). SRA has a more significant effect at low w/c as shown in figure 2.12.

Moreover, SRA was found to induce early expansion after the time of set, maintain a higher relative humidity level in the concrete, and to facilitate water loss from smaller pores, which results in reducing the concrete shrinkage as discussed by (Weiss et al., 2008). Greater early-age shrinkage was observed with the addition of superplasticizer (SP) as a result of improving cement dispersion, which consequently increases the rate of hydration reactions (Holt, 2005, Esping, 2007). In addition, it was pointed out that excessive SP dosage would delay the setting time and result in higher early-age drying shrinkage (Holt and Leivo, 2004). Scattered data on the effect of using air entraining admixtures on the early-age shrinkage were observed (Hammer and Fossa, 2006). However, the general trend was an increase of the autogenous shrinkage rate before it diverged from chemical shrinkage, and then it decreased thereafter. Conversely, air entraining admixtures were found by others to cause a considerable decrease in early age shrinkage (Kronlof et al., 1995).

Figure 2.12; Free shrinkage with different w/c and 1% shrinkage reducing admixture (D’Ambrosia, 2002).

e) Pozzolanic materials: the type, fineness, and percentage of cement replacement are the main parameters controlling the effect of pozzolanic materials on early-age shrinkage. Silica fume was found to increase the autogenous shrinkage significantly due to refining the pore structure of concrete (Tazawa, 1999, Wiegrink et al., 1996). Similar behavior was observed for ground granulated blast furnace slag (Lee et al., 2006, Lim and Wee, 2000). A high level of cement replacement by metakaolin (MK) (10-15%) was found to reduce both autogenous and drying shrinkage at early-age. This reduction may be a result of the dilution effect of reducing the cement content (Brooks and Megat-Johari, 2001, Kinuthia et al., 2000). Conversely, Gleize et al. (2007) observed an increase in autogenous shrinkage due to MK addition, which was interpreted as a result of the heterogeneous nucleation of hydration products on the surface of MK particles. Fly ash was also found to reduce the autogenous shrinkage of concrete (Lee et al., 2003, Termkhajornkit et al., 2005). In contrast, it was reported that very fine fly ash had a similar effect to that of silica fume (Tazawa, 1999).

f) Curing conditions: the curing method, duration and temperature have a significant effect on early-age shrinkage. Shrinkage was found to increase with increasing curing temperature during the first 5 to 10 hours after casting, and then it decreased with temperature increase. This may be attributed to an increase in moisture loss at high curing temperature, leading to an increase in the development of plastic shrinkage (Zhao, 1990). Additionally, moist curing with burlap and/or cotton mats was found to effectively reduce the early-age shrinkage and to increase water retention compared with other curing methods (Nassif et al., 2003, Huo and Wong, 2006). Furthermore, the longer the curing period, the lower was the shrinkage deformation (Tazawa, 1999).

2.10 Compensating for Early Age Shrinkage

Different methods have been developed to compensate for and/or reduce early-age shrinkage. Table 2.2 summarizes some of the different methods.

Table 2.2; Shrinkage Compensation Method

Methods

Effective Parameters

Actions

Reference

Low and moderate heat cement

High C2S content & formation of ettringite

Expansion

(Tazawa, 1999)

Shrinkage

reducing

admixture

Reduce surface tension of pore solution.

Reduce capillary

tension

(Bentur, 2003,

D’Ambrosia, 2002,

Holt, 2005, Esping,

2007, Lura et al.,

2007, Bentza et al.,

2001)

Fiber

High elastic modulus

Resist exceeded

tensile stress

(Wongtanakitcharoen

and Naaman, 2007,

Kronlof et al., 1995)

Reinforcement

schemes

Light reinforcement bars.

Resist exceeded

tensile stress

(Sule and van Breugel,

2001, Weiss, 1999)

Additive

Expansive additive,

gypsum, fly ash,

metakaolin, water repellent powder.

Hydration

product with

tendency to

volume increase

(Brooks and Megat‐

Johari, 2001, Tazawa,

1999, Kinuthia et al.,

2000)

Internal curing

Soaked lightweight agg

Water soluble,

chemicals (self‐curing

admixture), Smart

paraffin microcapsule,

Superabsorbent polymer.

Store water for

internal curing,

increase water

retention,

Mitigate rapid

Temperature change.

(Mihashi and Leite,

2004, Esping, 2007,

Duran‐Herrera et al.,

2007)

2.11 Creep

Creep of concrete can be defined as the permanent movement of a material in order to relieve stresses within the material. It occurs when the concrete is under load for a long period of time. To understand creep one can consider two identical specimens subjected to exactly the same environmental histories; one specimen being loaded and the other (companion specimen) load-free. Creep is commonly defined as the strain difference between a loaded and a companion load-free specimen.Creep can be divided into two, basic creep and drying creep. Basic creep occurs if the concrete is sealed or if there is no moisture exchange between the concrete and the ambient medium. Drying creep is the additional creep experienced when the concrete is allowed to dry while under sustained load. The sum of basic and drying creep is referred to as total creep.

The creep strain at any time Ec (t), is determined as:

Ec (t) = E(t) – Ee – Esh (t) Eq. 5

Where,

Ec (t) = creep strain at any time

E (t) = Total measured strain at any time t

Ee = Average instantaneous elastic recorded immediately after

loading

Esh(t) = drying shrinkage strain at any time (determined on unloaded

specimen).

The mechanism of early-age creep is not completely understood. Several mechanisms have been proposed including real and apparent mechanisms. Real mechanisms, involving the viscous flow theory, plastic flow, and seepage of gel water, are related to cement hydration and can be considered as material properties (Bentur, 2003). Apparent mechanisms are associated with micro-cracking and stress-induced shrinkage (Altoubat, 2002). Generally, creep is a combination of these two categories of mechanisms. Several mathematical formulas were proposed to predicate and model early age creep. However, the accuracy of these formulas in capturing and simulating the early-age behavior of concrete is still questionable and need more investigation (Springenschmid, 1998).

Creep is mainly affected by the mixture composition of concrete, loading age and duration, water migration, temperature, moisture conditions, and the stress level (Neville, 1996). The earlier the loading age, the higher the creep strain values due to the low modulus of elasticity of concrete (Bissonnette and Pigeon, 1995, Neville, 1996). Furthermore, no relationship between the applied stress and the resultant creep strain was found for specimens loaded at an age of 24 hours (Lennart O., 2003). Increasing the temperature was found to enhance the early-age creep rate, which is contrary to its effect at later ages (Neville, 1996). However, high temperature is expected to reduce the early age creep value as a result of accelerating the hydration process (Springenschmid, 1998). Moreover, the early-age creep of restrained concrete was found to be inversely related to the w/c (Altoubat, 2002).

Generally, mineral additives such as fly ash, metakaolin and slag were found to reduce the tensile creep and relaxation of concrete at early-age (Brooks and Megat-Johari, 2001, Pane and Hansen, 2002, Li et al., 2002), while silica fume showed an opposite trend (Kovler et al., 1999, Bissonnette and Pigeon, 1995, Pane and Hansen, 2002, Li et al., 2002). The spherical shape and smooth surface of un-reacted silica fume particles at early-age were suggested as a reason for this behaviour (Kovler et al., 1999).

2.11.1 Effects of Creep

Creep of concrete is both a desirable and undesirable phenomenon. On one hand it is desirable as it imparts a degree of necessary ductility to the concrete. On the other hand, creep is often responsible for excessive deflection at service loads which can results in the instability of arch, or shell structures, cracking creep buckling of long columns and loss of prestress. Frequently the detrimental results of creep are more damaging to non-load bearing components associated with the structure, such as window frames, cladding panels and partitions, than they are to structure itself (Davis and Alexander, 1992).

2.12 Mechanical Properties

Mechanical properties must be determined in order to be able to model the behavior of early age concrete. The compressive and tensile strengths determine whether failure will occur, while modulus of elasticity gives an estimate of the stresses, which are building up as a result of the volumetric dilations and the degree of restraint. Also Poisson's ratio must be known in order to make 2D and 3D generalizations suitable for finite element modeling.

2.12.1 Compressive Strength

The compressive strength is considered as a key property of concrete. It provides a general indication of concrete quality. The gain in compressive strength is typically rapid at early-age, and then becomes slower at later ages. This early rapid increase in strength is directly related to the increase in the calcium silicate hydrate (CSH) gel/space ratio (Neville, 1996). Compressive strength is influenced by several factors; most notably the w/c, type of cement, additives, and curing conditions (Lennart O., 2003). Rapid hydration of cement results in a higher degree of hydration and consequently higher early-age strength for a given w/c ratio (Mehta and Monteiro, 2006).

Curing conditions, i.e. the availability of moisture and the temperature profile, drastically affect the compressive strength gain. At a very early-age, the absence of moisture usually has a limited effect on the early strength gain, because concrete is still wet. However, inadequate moisture curing during the first day after casting concrete could lead to noticeable strength loss at later ages (RILEM, 1981). Furthermore, a high initial curing temperature speeds up the hydration reactions and the formation of the hydrated cement paste structure at early-age. Thus, it enhances the early-age compressive strength of concrete. However, it decreases the strength at later ages due to the lower quality of hydration products microstructure formed at higher temperature (Kahouadji et al., 1997).

Pozzolanic materials can also contribute to the early-age strength through improving the particle packing density (filler effect) and densifying the aggregate-cement paste interfacial transition zone (Neville, 1996). In addition, the type and level of addition of pozzlans have a significant influence on the compressive strength development. For instance, concrete incorporating class C (high calcium oxide) fly ash generally develops higher early-age strength than that of concrete made with class F (low calcium oxide) fly ash (Kosmatka et al., 2002). Moreover, other pozzolanic materials such as blast furnace slag were found to slightly improve the early-age strength compared to its more significant contribution to the later strength, which can be ascribed to its slow hydration rate (Shan-bin and Zhao-jia, 2002). However, applying the crystal seed technology, i.e. addition of hydrated micro-crystals, for this type of pozzolanic materials was found to improve its early strength contribution significantly (Prusinski, 2006). Generally, using pozzolanic materials may result in higher or lower early compressive strength of concrete depending on their type, addition rate, mineralogy, particle shape and fineness, and pozzolanic activity.

Several methods have been developed to predict the compressive strength of concrete at early-age. One of the well-known methods is the maturity or the "equivalent age" method, which is expressed as a function of the time and temperature of curing as follows:

M(t) = ∑ ( Ta – T0) ∙ ∆t Eq. 6

Where; ∆t is the time interval,

Ta is the average concrete temperature during the time interval

∆t, and T0 is the datum temperature (Mehta and Monteiro, 2006).

At early-ages, maturity is low, especially in the first two days after casting, and the relation between compressive strength and maturity is not fully linear. Thus, for accurate compressive strength prediction at early-age using the maturity concept, the heat of hydration should be considered during the calculation of maturity (Oluokun et al., 1990). Figure 2.13 shows the average additional maturity due to the heat of hydration with respect to concrete strength (σc).

Figure 2.13 Average additional maturity, (F-hr), due to hydration (Oluokun et al., 1990)].

2.12.2 Tensile Strength

The development of the tensile strength of concrete is of outmost importance for the prediction of the crack initiation. The uniaxial tension test which is believed to give the best estimate of the tensile strength is not widely used due to the difficulties in conducting the test. Instead, several indirect methods have been developed, e.g. the split tensile test (also known as the Brazilian test) and the three point bending test (which gives the modulus of rupture). However, the interpretation of these indirect test methods often relies on linear elastic formulas combined with correction factors determined empirically. This fact make the use of the methods unreliable if the tensile strength in unusual situations like concrete in early age or fibre reinforced concrete, are to be determined. This is due to the fact that the correction factors are compensating for the actual behaviour of the concrete, which is not linear elastic and ideal-brittle, but quasi-brittle. And the brittleness of the concrete change significantly in early age and fibre reinforced concrete compared with matured, normal strength and fibre-free concrete.

Experimental results based on the uniaxial tension test are few in the literature. This is caused by the difficulties of conducting the test, due to the problems with self-weight and frictional forces, which become significant in early age. Specimens that are tested in upright position are influenced by self-weight while specimens that lay down are influenced by friction against the sub-base. The latter may however be reduced by the use of teflon sheets. Furthermore, the results are often reported as function of the compressive strength or the splitting tensile strength. Although this seems relevant from a practical viewpoint it blurs the development of the uniaxial tensile strength since the behaviour of the other test methods change with brittleness and age (Lennart O., 2003).

The tensile strength of concrete can be related to its compressive strength. This relationship is influenced by age, grading, type and density of aggregates, curing conditions, and the strength evaluation method. At early-age, the tensile strength tends to increase more rapidly than the compressive strength (RILEM, 1981, Bentur, 2003, De Schutter and Taerwe, 1996). Conversely, some researcher found that the tensile strength increases at a lower rate than that of the compressive strength (Swaddiwudhipong et al., 2003), and that the ratio between the two properties decreases from 0.1 to 0.04 as the concrete matures. This discrepancy can be attributed to differences in the starting age of the test and quality of concrete. (Kasai et al., 1972) reported a higher increasing rate for the tensile strength compared to that of the compressive strength during the very early age up to around 0.5 day. Thereafter, the tensile strength increasing rate became lower than that of the compressive strength, as shown in Figure 2.14

Figure 2.14: Tensile / compressive strength of concrete versus age.

The tensile strength evolution between the ages of 2 and 8 hours after mixing concrete was investigated by (Abel and Hover, 1998). A dormant period of tensile strength gain was observed from the age of 2 hours to 4 hours at which tensile strength was infinitely low. This period was followed by a very rapid tensile strength development starting at the point of initial setting. However, it is difficult to monitor this dormant period for mixtures characterized by a delayed setting. On the other hand, the tensile strain capacity (strain at maximum stress) starts with a very high value before the initial setting and continues to decrease beyond that, reaching a minimum value later (figure 2.15) (Hammer et al., 2007).

Figure 2.15 Uniaxial tensile strength, and b) uniaxial tensile strain capacity

2.12.3 Modulus of Elasticity

According to Krauss and Rogalla (1996), the modulus of elasticity affects the stresses in the concrete more than any other property. The modulus of elasticity determines the conversion ratio of strain to stress in the concrete. As the strain is given for both shrinkage and thermal movement, a lower modulus of elasticity will decrease the stress in the concrete. However, a lower modulus of elasticity comes from a concrete with a lower binder ratio and thus usually lower the strength as well (Xi et al., 2003).

A concrete’s modulus of elasticity approximately mirrors the concrete’s strength. It is unclear if there is any net benefit from reducing the binder ratio, since the strength is usually reduced. Of course, the external loads apply a given stress to the system, so a lower modulus of elasticity will increase deflections except that the effect will simply be a reduction of the load taken by the deck and an increase of the load taken by the girders (whose modulus of elasticity is a constant).

To reduce the modulus of elasticity without reducing the strength, the primary approach is to use aggregates with a low modulus of elasticity (Xi et al., 2003; Krauss and Rogalla, 1996). Aulia (2002) also found that the modulus of elasticity was largely dependent on the aggregate used, and demonstrated that the relationship held true in fibre-reinforced concrete as well. Whether choosing aggregate to give a low modulus of elasticity is practicable depends on the location where the concrete is batched.

2.12.4 Poisson’s Ratio

Poisson’s ratio is the ratio between the lateral and longitudinal strains under the same uniaxial load (Neville, 1996). Limited research focused on the evolution of the Poisson’s ratio of concrete at early-ages. A review of previous studies showed that the Poisson’s ratio changes with time. It starts with a high value at early stage, then it decreases sharply until it reaches a low value during the first 24 hours, then it starts to increase again (Mesbah et al., 2002, Byfors, 1980). Conversely, other researchers stated that the Poisson’s ratio of concrete does not change significantly with age, and may be considered constant (Oluokun et al., 1991). Considering the assumption that the deformation of fresh concrete occurs without volume changes, the Poisson’s ratio can be taken equal to 0.5 (De Schutter and Taerwe, 1996). Table 2.3 shows a typical Poisson’s ratio values observed by different researchers (RILEM, 1981, Mesbah et al., 2002, Byfors, 1980, Oluokun et al., 1991). It was found that the Poisson's ratio does not change significantly with the increase of the cement content (Oluokun et al., 1991, Oluokun, 1989). Moreover, tests conducted on concrete specimens at different ages (1, 3, 7, 14, 28, 90 and 180 days) have shown an insignificant effect of the moisture content and ambient temperature on Poisson’s ratio (Downie, 2005).

Table 2.3 Poisson’s ratio values

Initial Value

Minimum Value*