Shrinkage of high performance lightweight concrete exposed to hot-dry weather conditions

The main aim of this investigation is to study the combined effect of hot-dry weather conditions on the plastic and drying shrinkage of high performance lightweight aggregate concrete (HPLWAC) specimens, along with the other properties including, workability, setting time compressive and splitting strength. The experimental program including the use of fixed mix proportions and was carried out in a typical Iraqi summer day (under actual conditions) of different times during the day. The results indicate that the use of lightweight aggregate play a main role in decreasing the effect of hot dry weather on plastic shrinkage, as well as the plastic shrinkage strain of HPLWC specimens cast at 12:00 pm. less than that of normal weight aggregate concrete specimens by about 36.7%. While drying shrinkage of HPLWC specimens at initial ages up to 7 days was low and it increases with a higher rate at later ages.


Introduction
High temperature, low RH and high wind speed alone or a combination of these factors encourage rapid evaporation of bleeding water that results in shrinkage, thereby inducing tensile stresses in concrete.If stresses induced due to shrinkage, are greater than the tensile strength of concrete, cracks are formed [1] [2].The use of high performance concrete in hot dry weather conditions normally requires special precautions against excessive water evaporation, because it has high-cement content and low-water to cementitious materials ratio, the low bleeding rates may result in early age cracking, which is associated with self-desiccation and autogenous shrinkage.In order to avoid such a risk of cracking, it is necessary to prevent the internal relative humidity from decreasing as the hydration of cement takes place [3] [4].
Due to the rapid development of very tall buildings, long span and large-size concrete structures, the requirements for better concrete performance are, higher strength, lightweight, higher toughness and others, therefore, high performance lightweight concrete (HPLWC) has been used for structural purposes.

Cement
Ordinary Portland cement (type 1) manufactured by Mass Bazian Factory was used in this study.Both the chemical composition and physical properties of the cement conformed to I.Q.S No. 5/1984 [5].

Silica fume (SF)
Silica fume used was commercially known as (Sika fume-HR), it was brought from Sika Company.The physical properties and chemical composition of the silica fume show that the silica fume used satisfies the requirements of ASTM C 1240-06 [6].

Fine aggregate
Natural sand brought from Al-Ukhaider region was used as fine aggregate passing through sieve 4.75 mm.The results demonstrate that the fine aggregate grading and the sulfate content conform to the I.Q.S No.45/1984 [7] Zone 2. Also, lightweight fine aggregate (Pumice) was used as a partial replacement by volume of the natural sand, this lightweight aggregate was crushed by crusher machine.The gradation of pumice fine aggregate was prepared as that of the natural sand.Specific gravity, absorption and dry loose -unit weight of pumice fine aggregate was 1.32, 17.3%, and 968 kg/m 3 , respectively.

Coarse aggregate
Natural crushed gravel of nominal maximum size of 12.5mm was used as coarse aggregate.It was brought from Al-Nebai quarry region.The grading of coarse aggregates met ASTM C33/03 requirement [8].Lightweight coarse aggregate (pumice) was also used as partial replacement of normal coarse aggregate.Grading of pumice coarse aggregate falls in the size designation of 12.5 to 2.36 mm in accordance to ASTM C 330-04 [9].The specific gravity, absorption and dry loose -unit weight of pumice coarse aggregate are 0.93, 28.6%, and 672 kg/m 3 , respectively.

High range water reducing admixture (Superplasticizer)
A high range water reducing agent was used (commercially known as Sika Visco Crete -5930), and meets the requirements for superplasticizer (Sp) according to ASTM C -494 type F [10].

Concrete Mixes
Mix design is carried out in accordance with volumetric method of ACI 211.2 [11].The reference concrete mixture is designed to have a characteristic compressive strength of 40 MPa at 28-days.Only one mix with 1: 1.33: 0.84 (cement: fine aggregate: coarse aggregate) by weight was used throughout this investigation.The design was carried out to conform to the requirements of structural lightweight aggregate concrete, according to ACI Committee 213 classification [12].
Concrete mixtures contained 572 kg/m 3 cement, 10% by weight of cement silica fume, superplasticizer (1.5% by weight of cement), and 50% lightweight fine aggregate was used as partial replacement to normal weight fine aggregate (sand), and 80% lightweight coarse aggregate used as partial replacement to crushed gravel.
Mixing process was carried out in the field under actual hot-dry weather conditions in a typical Iraqi summer days at different hours of the day at (7:30am, 9:30am, 12:00 pm, 3:00pm), concrete mixes were denoted by (M 7:30, M 9:30, M 12:00, M3:00), respectively.Before mixing, all concrete ingredients and mixer as well as molds were kept under sun radiation.Other specimens were prepared in shadow site at (12:00) pm.viz.(M 12:00 Sh) and others in laboratory denoted by (M Ref).
Another concrete specimen denoted by (M-N.W.A) was prepared with the same mix proportions mentioned above but it contains only normal weight fine and coarse aggregate which was cast at 12:00pm.under sun radiation.

Mixing Procedure
Before mixing, the pumice aggregates were pre-soaked for 30 minutes in water and used after removal of excess surface water by spin-drying for 10-15 minutes in a centrifuge to produce the aggregate particles in state saturated surface dry (SSD) condition.A pan mixer of 0.1 m 3 capacity was used.Coarse aggregate was added first to the drum followed by fine aggregate.All these materials were first mixed for 1 minute to get homogenous materials.The required cement and SF were mixed manually until a homogenous mixture was obtained, then they were added to the rotary drum.Two thirds of the water was added to the dry mixture and mixed for 1 minute.Finally the superplasticizer and one third of the remained water were stirred and added to the mixture.After that the mixture was mixed for 6 minutes.

Curing
All the specimens were demolded after 24 hours from casting and water cured for 13 days, and then they were left in the same conditions of casting until the time of test.

Experimental Tests
The following tests and measurements were conducted on prepared mixtures:

Workability
The workability of all concrete mixes was measured at different times by using slump test immediately after mixing in accordance with test method ASTM C-143 [13].

Setting time
According to the method described in ASTM C 403 [14], the initial and final setting time of concrete was determined, by means of penetration resistance measurements on mortar sieved from the concrete mixture.

Compressive strength
The compressive strength test was carried out on cubes of 100 mm according to BS.1881: part 116 [15].

Splitting tensile strength
The splitting tensile strength test was carried out on cylinders of 100 ×200 mm according to ASTM C 496 [16].

Length change
The test method involves the measurement of length change of HPLWAC specimens, which permits assessment of the potential for volumetric changes of concrete specimens.The wooden molds 100 ×100 ×400 mm were used for casting the specimens.After 24 hours from casting, the specimens were demolded and left in air without any curing for a period of 6 months.A stainless steel demec points for 150 mm were fixed at selected positions on the sides of the specimen to measure the length change by using mechanical extensometer.The accuracy of the dial gauge of the measuring device is 0.002 mm.

Plastic shrinkage strain and plastic shrinkage cracking
The concrete specimens utilized for evaluating the effect of hot weather exposure conditions on the plastic shrinkage strain and plastic shrinkage cracking were 1000 ×1000×50 mm.Wooden forms were used to cast the concrete specimens.The concrete mixing process was conducted in the site in actual hot weather conditions at different times of the day (9:30 am., 12:00 pm., and 3:00 pm.) and others were cast in shadow site at 12:00 pm., after casting the specimens were left outside and exposed to atmospheric conditions.To measure plastic shrinkage strain, two aluminum strips measuring 25×70×2 mm were placed on the two corresponding sides of the specimen as shown in Figure1.The movement of each strip, due to shrinkage, was recorded through two linear variable differential transducers (LVDTs) that were connected to a data acquisition system.Shrinkage displacements were recorded every 30 minutes during 24 hours [17].Also the concrete specimens were visually monitored for 6 hours to detect the onset of plastic shrinkage cracks.The lengths of cracks were recorded and average width was measured by using optical microscope.The relation between the workability of HPLWAC represented by the slump and the actual hot-dry weather conditions is showed in the Table 1, where the workability is decreased by high temperature and decreased relative humidity.It has been noted that the initial concrete temperature effects on the workability of concrete and the amount of water required for a given slump due to evaporation.

Initial and Final Setting Time of Concrete
The penetration strength as a function of time were plotted to determine the initial and final setting time of HPLWAC specimens.They were determined using penetration resistance values of 3.5 MPa and 27.6 MPa, respectively as illustrated in Figure 2 and Table2.It can be seen that both setting times of (M12:00) are less than other specimens.This is attributed to the high initial temperature of these specimens, it is about 48.5°C, however, (M 3:00) had higher initial temperature of 49.7°C, but the setting time is more than that of (M12:00).This may be it exposed to sun radiation for a less period.On the other hand, both the initial and final setting times are reduced with the rise in temperature and the reduction in RH which accelerates the rate of concrete stiffening (i.e.slump loss) and increase the rate of temperature rise inside the concrete.Also it can be seen from pervious study [18] that the initial and final setting time of normal weight concrete cast at 12:00 pm.under actual hot-dry weather conditions was less than that of current study.It is about (2:15) hrs.and (3:39) hrs., this may be attributed to the internal reservoir of water, which is stored in porous lightweight aggregate (pumice), that causes a delay the setting time of concrete.

Compressive Strength
The compressive strength results at age (3, 7, 28, 90 and 180) days of HPLWAC specimens that were cast at different hours of the day under actual hot-dry weather conditions are shown in Table 3.The results show that the percentage increase in compressive strength at age 7 days of specimens cast at (7:30am., 9:30 am, 12:00pm.and 3:00pm) was (5.2, 6.5, 6.8, 6.2)%, respectively, compared with the 3-day strength, while for (M Ref) specimens the percentage increase is 4.6%.Whereas the percentage increase in 28day strength of specimens M (7:30, 9:30, 12:00, and 3:00) compared with the 7-day strength, is about (21.3, 23.3, 23.8, and 22.9)%, respectively, while for (M Ref) is 37.3%, also for the same specimens, the 90-day strength increased by about (8.7, 8.2 ,7.9, and 8.6)% compared to 28-day strength, whereas (M Ref) the percentage increase is higher, it is about 15.3%.The results also have shown that the percentage of increase in strength during 2 months is less than that during 21 days.These values are typically less than that for normal weight concrete.The strength development in lightweight concrete is limited by the inherent strength of aggregate.These results are in agreement with those found by Al-Khaiat and Haque [19].Also the percentage increase in 180-day strength compared to the 90-day is (5.1, 4.5, 4.4, and 4.7)% and for (M Ref) is 5.3%, this means that the gain in strength at 28 day and later ages for (M Ref) is higher than that for specimens cast and cured under hot weather conditions.This is in agreement with other researchers [1,20,21] which reported that "Concretes mixed, placed, and cured at elevated temperatures normally develop higher early strengths than concrete produced and cured at lower temperatures, but strengths are generally lower at 28 days and later ages"

Splitting Tensile Strength
The relationship between splitting tensile strength and age of HPLWAC specimens that were cast at different hours under actual hot weather conditions is shown in Figure 3.
Generally the results indicate that the pattern of the behavior of all specimens in splitting tensile strength is similar to that of the compressive strength.The figure display also that the splitting tensile strength at early ages (3, and 7) days of specimens M (7:30, 9:30, 12:00, and 3:00) was higher than that of reference specimens, the percentage increase at 3 days was (7.9, 13.2, 17.3, and 9.8) % respectively, and at 7 days the percentage increase was (9.6, 16.3, 20.6, and 11.7) % respectively, compared to reference specimens.On the other hand, the splitting tensile strength of specimens (MRef) at 180 day age was higher than that of specimens (M 7:30, M 9:30, M 12:00, and M 3:00) by about (11.8, 6.3, 3.9, and 10.6)%, respectively.

Length Change of HPLWAC Specimens
The data in Table 4 and Figure 4 indicate that drying shrinkage up to 7 days for specimens casted at (7:30 am., 9:30am., 12:00 pm., and 3:00 pm.) is low, and not exceeding (108, 92, 59, and 162) micro strain respectively.This behavior may be attributed to the process of internal curing in pumice concrete due to the supply of absorbed water from (pre-soaked pumice) aggregate pores into the finer capillary pores of cement paste that can reduce the early age drying shrinkage of concrete.Also, the evaporation process is reduced due to the formation of dense hardened paste and the segmented capillary pores.Because of this, the core of the prisms resists the surface contraction on drying, sometimes resulting in surface shell drying of the specimens [22].
From Fig. 4 it can be seen also that short-time exposure of fresh concrete is expected to reduce shrinkage, such reduction was observed in concrete specimens cast at 3:00 pm. which was exposed to intensive drying for a short time, maximum (peak) shrinkage was 616 micro strain at age 56 days, while peak shrinkage for specimens cast at (12:00 pm.) was higher than that of specimens (M-3:00) it is about 902 micro strain.Whereas in specimens M (7:30, and 9:30) the shrinkage at 56 days is less than that of specimens M (12:00, and 3:00), it is (102, and 300) micro strain.The effect of exposure time on shrinkage is due to the effect of drying on the structure of the concrete which, in turn, affects differently on the strength and shrinkage.While at a later age after 84 days, it can be seen that the shrinkage reduces for specimens M (12:00, and 3:00), it is of magnitude (524, and 410) micro strain.This may be due to the fact that there are no further volume changes in concrete, and the internal cracking occurs, the presence of cracks, including internal crack, reduces shrinkage because some of the induced strains are taken up by the cracks and are not reflected, hence, there is a reduction in measured shrinkage [23].While the peak shrinkage of specimens cast at (9:30 am., and 7:30 am.) was (735, and 767) micro strain, respectively at age 168 days, and for specimens cast in shadow site at 12:00 pm. was (670) micro strain at age 126 day.
On the other hand, it can be noted also from the figure that the maximum swelling occurred to the specimens during 180 days of exposure period of (200, 340, and 359) micro strain for specimens cast at (7:30 am., 9:30am., and 12:00pm) at age (70) days, respectively, and about 464 micro strain for specimens (M-3:00) which happened at age 98 days.This is because of the sudden changes in relative humidity due to the rain storms which results in an increase in the volume of the paste due to a rewetting of the specimens (i.e.swelling).

Plastic Shrinkage Strain and Plastic Shrinkage Cracking
The effect of actual hot dry weather conditions on plastic shrinkage strain of HPLWAC specimens is depicted in Figure5, which illustrates the variation of plastic shrinkage strain over a period of 24 hrs.for different cast times of specimens.It can be observed from the figure that the plastic shrinkage strain increases with the period of exposure to ambient conditions, and it increased almost approximately linearly with time up to about 4 hrs., after that the rate is decreased.It can be noticed also from Figure5 that the maximum plastic shrinkage strain is for specimens (M3:00), and around 809 micro strain after 24 hrs.from casting.This is because of the excessive evaporation of mixing water due to the high ambient air temperature, and low RH at the time of cast; they are about 58°C, and 10% respectively, leads to an initial concrete temperature of 46°C.This refers to the point that increased rate of evaporation and mix concrete temperature lead to excessive tendency for plastic shrinkage and cracking.From Table 5, it can be seen that the number of hair cracks of specimens (M-3:00) are more than those of other specimens.8 cracks with width ranged from 0.1 to 0.65 mm can be observed.While maximum plastic shrinkage strain for specimen cast at 12:00 pm. is 656 µm over a period of 24 hrs.from cast, which is about is 19% less than that of specimen cast at 3:00 pm.This may be due to the fact that initial concrete temperature of specimen cast at12:00 pm. was less about 44°C, since air temperature was 55°C and RH was 13%, and the specimens (M-12:00) have less number of hair cracks than those of the specimens (M-3:00), which are 4 cracks, with width ranging from 0.06 to 0.7 mm.Whereas specimen cast at 9:30 am. has a maximum plastic shrinkage strain of 463 micro strain, that is (43 and 29.5) % less than those of specimens cast at (3:00 and 12:00) pm., respectively.This may be because the air temperature at time of cast specimens (M-9:30) is low at about 42°C and the RH is 17% and the initial concrete temperature was 38°C, this means that the plastic shrinkage strain decreases with reducing temperature and increasing RH, also plastic shrinkage cracks of specimens (M-9:30) are short ranged from 5 to 10 cm, and with a small number of hair cracks nearly 4 cracks.The results of the maximum plastic shrinkage strain of current work are in agreement with the findings of Al-Amoudi et al. [17].12:00 pm.
12:00 pm. Figure 6 shows that specimen (M12:00Sh) has plastic shrinkage strain at 24 hrs.less than that of the specimens cast under sun radiation by 18.5 %.The increase plastic shrinkage strain for concrete specimen (M-12:00) may be attributed to the low bleeding [17].The absence of bleed water in the fresh concrete enhances the evaporation of pore water from surface concrete layers under hot dry weather.The evaporation of mixing water from fresh concrete induces shrinkage stress in it [17].

Air
It can be seen also from Figure 6 that the plastic shrinkage strain of specimens (M-N.W.A) is 1037 µm at 24 hrs., which is higher than that of specimen (M-12:00) by 58%.This may be attributed to the fact that the lightweight concrete specimen had low modulus of elasticity about 24.4 GPa [24].The lower the modulus of elasticity, the lower will be the amount of the induced elastic tensile stress for a given magnitude of shrinkage [20].

Conclusions
The following conclusions can be drawn from the results and discussions presented from this investigation: 1.Using pre-soaked lightweight aggregates (pumice) as internal water reservoirs for producing HPLWAC under actual hot-dry weather condition played positive role in improvement of concrete properties by compensating the evaporation of water due to the rising temperature and decreasing relative humidity, and provide additional moisture in concrete for a more effective hydration of the cement.
2.The compressive strength and splitting tensile strength performance results of specimens (M 12:00) cast and cured under actual hot-dry weather and exposed to sun radiation was more than that for specimens of shadow site at all ages.While strength performance of reference specimens was slightly more than that of specimens (M 12:00) at later ages.
3.The plastic shrinkage strain of HPLWAC specimens increases almost linearly with time up to about 4 hours, after that the rate of increase in plastic shrinkage decreases until 24 hours.The highest plastic shrinkage strain was 809 micro m in the concrete slab cast at 3:00 pm.under hot dry weather conditions.While the lowest plastic shrinkage strain was 463 µm in concrete slab cast at 9:30 am.Also plastic shrinkage strain of HPLWAC specimens cast at 12:00 pm. is less than that of normal weight aggregate concrete specimens by about 36.7%.4.The drying shrinkage of HPLWAC specimens at initial ages up to 7 days is low.It does not exceed (108, 92, 59, and 162) micro strain for specimens cast at (7:30 am., 9:30am., 12:00 pm. and 3:00pm.),respectively.Drying shrinkage increases with a higher rate at later ages.Maximum drying shrinkage for specimens M1 (7:30, and 9:30) at age 168 days was (767, and 735) micro strain respectively, while for specimens M1 (12:00, and 3:00) was (902, and 616) micro strain respectively at age of 56 days.Whereas, specimens cast in shadow site at 12:00 pm. was 670 micro strain at age 126 days.

Fig. 2 .
Fig. 2. The relationship between penetration resistance and time of HPLWAC specimens

Fig. 3 .
Fig. 3. Relationship between splitting tensile strength and age of HPLWAC specimens casted at different hours under actual hot-dry weather conditions

Fig. 4 .
Fig. 4. Length change of HPLWAC specimens cast at different hours in hot dry weather conditions

Fig. 5 .
Fig. 5.The effect of actual hot dry weather conditions on plastic shrinkage strain of HPLWAC cast specimens at different hours

Table 5 .
Plastic shrinkage cracks of HPLWAC specimens cast at different hours in actual hot weather conditions

Fig. 6 .
Fig.6.The variation of plastic shrinkage strain of concrete specimens cast at 12:00 pm.

Table 1 .
Slump of HPLWAC specimens cast at different hours of the day in hot dry weather conditions

Table 2 .
Initial and final setting time of concrete specimens cast at different hours

Table 3 .
Compressive strength of HPLWAC specimens cast in hot dry weather conditions

Table 4 .
Length change of HPLWAC specimens cast under hot dry weather conditions