The tensile deformation and capillary pressure build up in fresh concrete

. During the plastic state of concrete, any hindrance or resistance of the free volume change in plastic concrete induce tensile stresses and or strains in the concrete element. Crack formation is expected to occur if the tensile stress and strain is greater than the capacity of the concrete. Investigation into the tensile properties and relaxation behaviour of plastic concrete was carried out using a direct tensile testing machine. The capillary pressure was measured during the tensile tests in low evaporation conditions, as well as in a climate controlled chamber where the concrete was exposed to high evaporation conditions. Most of the measured strength gain (tensile capacity) of the concrete is due to the capillary pressure in the pores of the fresh concrete which keeps the particles together by means of free water in the concrete during the early stiffening phase of the concrete. Later the hydration products bridge the pores which provides strength to the concrete. The capillary pressure results indicate how the rate of hydration influence the interconnectivity of the pores, and the contribution to the measured strength gain of the fresh concrete. The capillary pressure measurements during tensile tests revealed that the mechanism behind relaxation is the negative capillary pressure build-up induced by the mechanical tensile strain. The results also showed a correlation between the build-up of the capillary pressure in the concrete and the tensile deformation of the fresh concrete where the capillary pressure increased as the tensile load increased .


Introduction
Fresh concrete, or concrete in its plastic state, typically refers to concrete in the first few hours after the concrete has been batched and cast [1].Initially, fresh concrete has very little tensile strength or stiffness.The concrete only develops strength as the hydration reaction of the cement in the concrete starts and binds the concrete particles [2][3][4].Prior to this strength development, fresh concrete is more susceptible to volume change.The volume change in fresh or plastic concrete is due to plastic shrinkage and plastic settlement.Plastic settlement is a onedimensional vertical volume change, while shrinkage is a three-dimensional volume change [5,6].
Any resistance or hindrances in the concrete will restrain the concrete and prevent free volume change.Such restraints typically occur in the form of reinforcing bars and/or sectional dimension changes in the concrete [1,5].Where the fresh concrete is restrained from free volume change, tensile stresses and or strains could develop in the concrete [4,5,7].
Plastic shrinkage is mostly due to the removal of water from the capillary pore structure of the concrete.As the pore water continues to be drawn from the concrete, a negative capillary pressure builds up in the pore structure of the concrete [8][9][10].This represents an internal loading aspect within the concrete as the concrete dries out and internal stresses occur.Applying curing procedures, which decrease the water loss from the concrete, reduces the build-up of capillary pressure.This is indicative of the unloading aspect of the concrete element as the capillary pressure decrease.The behaviour of loading and unloading can be linked to the relaxation behaviour of concrete.
During the plastic state of concrete, the concrete goes through various phases known as the stiffening, setting and hardening phases [11][12][13][14].Fig. 1 show these hydration phases corresponding to the concrete's state of matter.
The stiffening phase marks the first phase of cement hydration and starts as soon as water is added to the dry constituents.This phase can be defined as the loss of consistency of the plastic cement paste [12].During this phase a short dormant period also occurs where the reaction between the cement and water is inactive [1].
The second phase is called the setting phase and refers to the solidification of the plastic cement paste.During this phase the transition from a plastic material to a solid material occurs which is due to a decrease in the permeability of the concrete paste.The permeability refers to the rate of viscous flow of fluid under pressure through the pore structure [13].
The final hardening phase is where the concrete starts to gain significant strength.At the beginning of this phase the cement paste possesses nearly no strength since the hydration of the silicate Alite is still in the beginning stages [12].The continued hydration results in the progressive filling of voids with calcium silicate hydrate crystals, which decreases the porosity and increases the strength of the hardened cement paste.[11][12][13][14]).

Fig. 1. Phases of concrete hydration (adapted from
During the plastic phase, tensile stresses occur in the concrete when the concrete deforms due to settlement or shrinkage.Where the concrete has not gained sufficient tensile strength to resist these tensile stresses, cracks could form [5,16].
As the pore water gets extracted from the concrete, a negative capillary pressure builds up in the concrete.As more pore water is extracted from the concrete the curvature of the water menisci increases until a point where the water menisci cannot bridge the gap between the particles, resulting in the entrance of air into the pore system which results in a breakthrough in the capillary pressure [8][9][10].
Capillary pressure is the main mechanism responsible for plastic shrinkage cracking, while the tensile strength of the concrete determines if cracking is going to occur.Yet no references could be found where the capillary pressure build up is measured while testing the mechanical tensile strength of the concrete.Therefore, this study set out to investigate the tensile behaviour of plastic concrete and the role of capillary pressure build up during this plastic period.The relaxation behaviour of plastic concrete as well as its resistance to multiple loadings was also investigated.The experiments performed to achieve these objectives as well as the results achieved are discussed in the following sections.

Tensile test setup
The tensile test machine is made up of a support beam, fixed and free loading platforms, a mechanical actuator and an air bearing as shown in Fig. 2. The tensile mould comprises of two halves and was used to conduct all tensile tests.

Fig. 2. Tensile testing machine.
The mechanical actuator is directly bolted onto the support beam and provides the horizontal displacement required to induce a tensile force in the concrete filled mould.The actuator is capable of applying a maximum tensile load of 1.7 kN.This load is sufficient to induce cracking in conventional concrete during the plastic state.
A loading rate of 0.25 mm/min was used for all tensile tests as a loading rate that is not too fast or too slow [17].The tensile properties of plastic concrete changes significantly during the early ages of concrete and therefore too long testing periods do not reflect the tensile properties of the concrete specimen at a specific age of testing.A too fast loading rate does provide a quicker testing period but it increases the difficulty in capturing the full tensile behaviour of the plastic concrete [17].
The air bearing allows for frictionless movement while the moving platform attached to the actuator pulls the two mould haves apart.The air bearing achieves this by creating an ultra-thin film of highly pressurised air between the top surface of the air bearing and the bottom surface of the aluminium mould.This ultra-thin layer of air supports the concrete filled mould by elevating the entire mould by a small distance, limiting the contact between the air bearing and the mould which effectively reduces friction to a near zero magnitude.
The mould used to conduct the tensile tests on the concrete samples is shown in Fig. 3.The mould comprised of a dog-bone shape, machined from aluminium using CNC technology.The mould was constructed of aluminium to keep the weight of the mould as low as possible.The dogbone shaped mould contained two curved transitions that reduce the surface area of the concrete within the gauge area.The first function is to help grip the concrete and effectively transmit the induced tensile load to the concrete as the two mould halves move apart.The second function is to minimise any stress concentrations occurring outside the gauge area.This effectively limited the cracking to the gauge area of the mould.Each mould comprised of a moving and a stationary halve.An oil lubricant was lightly sprayed over the mould surface to decrease friction and to release the concrete from the mould.
A load cell located on the fixed loading platform measured the corresponding force induced within the concrete as the moving platform induces a tensile load on the concrete specimen.Two LVDT's on either side of the concrete filled mould is used to measure the displacement of the mould and a single LVDT is fixed to the centre of the mould captured the displacement of the concrete.Two internal pressure measuring sensors are inserted into the gauge and out-of-gauge zones within the concrete to measure the build-up of capillary pressure in the concrete.The measuring of the capillary pressure followed a similar method to that used by Slowik et al. [18].Data acquisition was set to 50 Hz which captured 50 data points per second.The measuring setup is also shown in Fig. 4.

Tensile properties tests
The objective of these tests were to obtain the development of the tensile properties of plastic concrete with time.Tests were carried out at 60, 120, 180 and 240 minutes using the tensile testing setup.At least 4 concrete samples were tested until well after failure at each of the four intervals.

Relaxation tests under single pre-peak loading
These tests were conducted to investigate the relaxation behaviour of concrete when subjected to a single pre-peak loading.Once the mould was set up on the air bearing and ready for testing, the concrete was loaded to 50% of the average maximum tensile strength for each corresponding hourly test followed by an abrupt stop of loading by ceasing the displacement of the actuator.Relaxation of the induced tensile force was then captured for a period of 10 minutes.At least 2 concrete samples were tested at each of the four intervals.

Relaxation tests under multiple loading cycles
The relaxation tests were repeated and allowed to relax for a time period of 5 minutes before loading the concrete again to 50% of the average maximum tensile strength in the tensile regime.This process was repeated at least three times or until cracking occurred.At least 2 concrete samples were tested at each of the four intervals.

Environmental simulation
After mixing and placement of the concrete in the mould, all samples were cured in a climate chamber until the time of testing.The climate chamber was set to an ambient temperature of 40ºC and wind speed of 4 km/h.These climate conditions were chosen to ensure curing under high temperature to increase setting times in order to shorten the plastic period of concrete, which then minimises the number of tests required.Furthermore, the samples were continuously sprayed with water to ensure that sufficient water remained on the surface of the samples.This prevented evaporation of concrete pore water and therefor any plastic shrinkage which can cause unwanted stresses in the concrete before testing.

Concrete mix
A standard conventional concrete mix with a design strength of 50 MPa, a slump value of 70 mm and a 0.55 w/c ratio was used throughout the study.A Greywacke stone with an angular particle shape and nominal size of 6 mm was used.A Siliceous natural quarry sand known as Malmesbury sand was used.The sand is formed through natural disintegration of rock and has a rounded particle shape with a fineness modulus of 1.39.The material constituents, proportions and properties are summarised in Table 1.

Moving mould halve Stationary mould halve
The setting times of the concrete mix was also determined.The initial setting time for the mix was recorded at 135 min whilst the final setting time was at 195 min.

Tensile strength
The strength development with time graph in Fig. 5 was plotted using the peak stress at point of cracking for all test periods, at 60, 120, 180 and 240 minutes.Peak stress values were averaged and plotted with the corresponding error bars representing one standard deviation higher and lower than the average.These results show the tensile strength development during each phase of concrete, namely stiffening, setting and hardening phases and was plotted against the three hydration phases in Fig. 5.
The 60 minute test samples showed very little strength development from time of placing to testing.Towards the end of the stiffening phase, at the 120 minute test samples, the hydration reactions start taking effect as can be seen from the significant strength gain from 120 minutes onwards.

Fig. 5. Strength development with time.
The initial setting time occurs towards the end of the stiffening phase which further marks the end of the window for placing and consolidating the concrete.Once initial set is reached, the setting phase starts and the cement paste becomes stiff and unworkable due to the ongoing hydration reaction.The results show that between the initial and final setting times, the tensile strength developed rapidly which resembles the stiffening of the concrete paste.During this phase bleeding typically has ceased since the stiffening of the cement particles prevents settlement.
Shrinkage due to evaporation poses great risk for cracks to develop during the setting stage if proper care is not taken in keeping the concrete surface moist.The still low tensile strength of the 180 minute test samples demonstrates clearly why concrete is so vulnerable to cracking at this age.
The final setting time marks the beginning of the hardening phase and is where concrete starts to gain significant strength.Fig. 5 clearly shows an increase in gradient from the 180 to 240 minute test samples.
The concrete has now gained sufficient strength in tension and is therefore less prone to plastic shrinkage cracking.

Influence of capillary pressure on tensile behaviour
These experiments investigated the build-up of stress in water filled pores when a tensile load is applied.Fig. 6 shows the capillary pressure readings for both gauge and out-of-gauge pressure zones plotted against the stress versus strain results for time periods 60, 120, 180 and 240 minutes.The capillary pressure in the gauge region (referred to as gauge pressure) of the 60 minute results increased at a similar rate and showed a similar shape compared to the mechanical stress versus strain results of the concrete.The out-of-gauge pressure however, increased at a much slower rate.The results indicate that much of the strength gain observed for the 60 minute results was due to the capillary force in the pores of the fresh concrete which keeps the particles together.This suggests that the presence of water in plastic/fresh concrete was responsible for much of the tensile strength experienced for the 60 minute test results.
The gauge pressure for the 120 minute results increased at a much slower rate.The results show that the concrete has already started gaining some tensile strength from the hydration products; however, much of the strength measured was still due to the pressure of free water holding the solid particles intact.The out-of-gauge pressure results showed very little pressure build-up compared to the gauge area results.At this age the concrete was nearing the start of the setting stage and the concrete was already starting to gain tensile strength due to the commencement of the hydration reactions.This indicates that the deformation of concrete in the gauge area has very little influence on the rest of the concrete out of the gauge area.As the strength develops, the hydration products bridge the gaps between the surrounding molecules, reducing the interconnectivity of the pore water and therefore preventing the free movement of water.
The 180 minute results shows that the pressure build-up in the gauge area failed which marks the point of air entry at the pressure sensor.At this point, the concrete's stress continued to build-up until cracking or failure.The presence of free water is still believed to provide the concrete with some strength, however the tensile strength is now mainly from the hydration products.The out-of-gauge pressures showed no corresponding peak.
The 240 minute results showed a very low increase in gauge pressure as the tensile load was applied.This gives an indication of the low amount of free water in the concrete and the density of the hydration products around the pressure sensor tip.The low localised pore water pressure in the gauge area therefore indicates that the pores around the sensor tip were much more rigid, preventing the pores and pore water from moving and in turn causing an increase in gauge pressure.The out-of-gauge pressure continued to increase at a constant rate, however it is believed that this is due to ongoing evaporation.

Development of capillary pressure with time
Fig. 7 illustrates a simple representation of a more complex capillary pore system in plastic concrete.Each section aims to explain the negative capillary pressure experienced between particles in the gauge and out-of-gauge zones when a tensile force is applied to the pore system.

Fig. 7. Influence of capillary pressure on internal concrete particles.
The 60 minute results showed very little strength gain from time of placement to testing.The pore water diagram shows three particles surrounded by pore water with little to no hydration products between the pores.There is a slight water meniscus between the particles indicating that the pore water bonds are in tension as they hold the particles together.As the actuator induces a mechanical tensile force between Particles 1 and 2 in the gauge area for the 60 minute test, Particle 1 moves away from Particle 2 and as the pore pressure tries to keep the two particles in place, an increase in pore pressure is measured in the gauge area.The displacement of Particle 1 in the positive U direction causes Particle 2 to move by a smaller distance of U2, which in turn indices a smaller buildup in pore pressure between Particles 2 and 3.This process continues until a crack forms between Particles 1 and 2.
The ongoing evaporation at 120 minutes results in a greater pore pressure experienced between particles and is represented by the deeper concave meniscus formed between hydration products.At this age, hydration reactions are constantly filling the pores with hydration products.As a mechanical tensile load is applied to the pore system, the gauge pressure continues to build-up until a drop in pore pressure occurs, at which point the measured tensile strength of the concrete is believed to be due to the hydration products.Before the pore pressure drops, the pore pressure is believed to provide the concrete with most of its strength.The pressure build-up in the out-ofgauge area remained largely constant after the initial peak.This constant measurement is believed to be due to the hydration products which fill up the empty pores in the concrete, which also adds rigidity to the concrete and prevents Particle 2 from displacing back to its near original position.
After 180 minutes there is an increased pore pressure experienced between pores as ongoing hydration continues to fill up the empty pores with hydration products, in turn reducing the size of the pores, resulting in smaller menisci forming between hydration products as seen in the 180 minute section in Fig. 7.The hydration products continue to bridge the gaps in and between the empty pores and provide the concrete with most of its tensile strength.As a mechanical tensile load is applied to the concrete, Particles 1 and 2 move further away from each other.This distance however is much smaller compared to the 60 and 120 minute particle displacements.The smaller amount of free water between pores and smaller meniscus means that more tensile force is needed to break the pore water bonds.The gauge pressure measures an increase in pore pressure at this point.The increase in tensile load however, has no effect on the out-of-gauge pressure since the hydration products have already bridged the gaps between the pores, meaning that the out-of-gauge pores are held in place by the hydration products.
At 240 minutes a smaller meniscus form between the hydration products, indicating that the amount of free water in the concrete is significantly less compared to the 180 minute results.

Relaxation behaviour under single loading
The term 'relaxation potential' is used to quantify the relaxation results and refers to the initial drop in stress from the point of constant strain to the lowest stress value resulting due to the relaxation behaviour of the concrete specimen.The relaxation potential is therefore represented as a percentage of the total Both the 60 and 120 minute specimens dropped into the compression zone once the displacement of the actuator was ceased, as shown in Fig. 8.By ceasing the displacement of the actuator, a constant strain was applied throughout the test period.The corresponding drop in stress, as a result of the constant strain, represents the relaxation behaviour of the concrete specimen.The 60 minute specimens displayed an average relaxation potential of 307%, while the corresponding 120 minute specimens displayed a much lower 175% relaxation potential.

Fig. 8. Relaxation of concrete at 60 and 120 minutes.
Fig. 9 shows the relaxation behaviour of the 180 and 240 minute samples.The 180 minute sample displayed an average relaxation potential of 55%.During this time period, the concrete is nearing the end of the setting period and as a result, the concrete is much more rigid compared to the younger specimens.The 240 minute sample showed nearly no relaxation with an average relaxation potential of 6.5%.Both the 180 and 240 minute samples displayed an increase in gradient after the initial relaxation drop.

Relaxation behaviour under multiple loading
During the early ages of concrete, typically during the stiffening and setting phases, the concrete is likely to undergo multiple loading and unloading cycles if curing is applied.For example in high evaporation conditions, the concrete element experiences a stress build-up and by lightly wetting the atmosphere above the concrete's surface reduces this stress build-up.
Fig. 10 displays the relaxation behaviour of the concrete under multiple loading cycles at 60 and 120 minutes.At least two sets of multiple loading tests were carried out for all time zones.Both test periods relaxed into the compression zone, relieving the mechanically induced tensile stress.As each cycle progresses, the concrete sample was able to relax all of the induced mechanical stress, however a slight decrease in relaxation potential was observed as each cycle was completed.The main reason for the low stress needed to crack the concrete specimen, is possibly due to the cumulative amounts of strain induced during the multiple loadings.The 180 minute concrete was nearing the end of the setting phase and the results show that the concrete was not able to relax all the induced mechanical stress.As a result, the remaining tensile stress after the initial relaxation during the first cycle was carried over to the next cycle, where the total load induced within the concrete sample is equivalent to 50% of the average maximum tensile strength, plus the remaining stress of the previous cycle.The 240 minute sample was only able to complete one cycle before failure.The main reason for this occurrence is believed to be due to the brittle nature of concrete, which lacks the ability to relax stress build-up compared to the 60 and 120 samples.The small amount of stress relaxed during the first cycle, was not sufficient in allowing for a second successful loading cycle.The ability to complete multiple cycles such as for the 60 and 120 samples, gives an indication of the ductility and forgiving nature of the fresh concrete under stress build-up and deformation.At these ages, the concrete is still plastic nearing the start of the setting phase and results show that concrete is able to withstand multiple loading cycles before failure.
The 180 and 240 minute test samples fall within the setting and hardening phases of the concrete.Concrete at this age is semi plastic, beginning to solidify and the low amount of completed cycles displays the brittle nature of the concrete.Plastic shrinkage cracking is more severe during the setting period of the concrete.Furthermore, the 180 and 240 minute samples had lower strain capacities compared to the 60 and 120 minute samples.This possibly explains why the concrete at 180 and 240 minutes lacked the ability to withstand or resist multiple cycles compared to the 60 and 120 minute concrete.

Influence of capillary pressure on the relaxation behaviour of concrete
Fig. 12 shows the typical relaxation behaviour of plastic concrete and capillary pressure measured in the gauge and out-of-gauge zones of the 60 minute sample under single loading.Results show that as the actuator was loaded to the desired force, the capillary pressure in the gauge area increased at a similar rate to the peak value, similar to the concrete's mechanically induced stress.The out-of-gauge pressure increased at a slightly slower rate compared to the build-up of the gauge pressure and concrete stress before reaching a peak.Once the actuator was ceased, both the capillary pressure in the gauge zone and stress in the concrete, decreased in magnitude from the tension zone into the compression zone.The drop in stress recorded during the periods of no loading, was mainly due to the relaxation in capillary pores.The out-of-gauge pressure relaxes very slightly during points of no loading.This is seen in Fig. 13 by the sudden drop in stress and capillary pressure as the two centre particles move towards each other.This is followed by an increase in the radius of the water meniscus between Particles 2 and 3. Therefore the drop in stress experienced in both the gauge pressure and concrete stress, marks the rearrangement of internal particles in order to relieve the load induced during loading.The relaxation behaviour of concrete before the setting phase is therefore due to the rearrangement of particles surrounded by capillary pores in order to relieve the stress build-up induced during loading.The results show that as the actuator was loaded for the first cycle, the gauge pressure increased at a similar rate compared to the stress.Once the displacement was ceased, the stress of the concrete and gauge pressure relaxed into the compression zone, while the out-of-gauge pressure decreased at a much lower rate.The change in relaxation potential of the stress in the concrete decreased after each cycle to such an extent that the range of relaxation during the last successful cycle falls within the tensile regime.This change is also apparent in the gauge pressure readings and is believed to be the cause of the decrease in relaxation potential observed by the stress in the concrete.Therefore results indicate that the potential for relaxation in concrete corresponds to the potential for relaxation of the capillary pressure experienced within pores.
Fig. 15 displays the relaxation behaviour and capillary pressure under single loading at 180 minutes.As the concrete was loaded to 50% of its average tensile strength, the gauge pressure increased at a similar rate but to a much lower magnitude, while the out-of-gauge pressure displayed no corresponding effect.Concrete at this age displayed a reduced amount of relaxation compared to the younger concrete.This is believed to be the reason behind the reduced relaxation in concrete at these ages.The presence of hydration products and reduced amount of free water hinders the rearrangement of particles and therefore results in a lower amount of relaxation during points of no loading.As the actuator was loaded for a second time to initiate cracking, the gauge pressure increased at a similar rate compared to the concrete stress.This confirmed that part of the strength experienced under tensile failure was due to the capillary pressure within pores.Obtaining capillary pressure readings at 240 minutes proved extremely difficult due to the low amount of free water available within the pores.At 240 minute the concrete displayed a low relaxation potential.As the actuator was loaded, an instant break in pore pressure was observed which gives an indication of how dense the internal particles are arranged at 240 minutes.

Conclusions
This study investigated the tensile and relaxation behaviour of plastic concrete as well as its resistance to multiple loading.The significant conclusions that can be drawn from this study are discussed.
• The capillary pressured tests show that most of the measured strength gain during the stiffening phase of the concrete is due to the capillary pressure in pores of the fresh concrete which keeps the particles together.This therefore indicates that the presence of free water in plastic concrete in responsible for much of the tensile strength experienced for the 60 and 120 minute concrete.Although the 180 minute concrete displayed a higher tensile strength than the younger concrete, much of the strength measured is believed to be due to the hydration products bridging the gaps between the molecules.However, the presence of free water in the gauge area indicates that part of the measured strength is also due to the capillary pressure holding the particles in place.• The capillary pressure results give an indication of two phenomena.The first describes the rate of hydration affecting the interconnectivity of pores, while the second describes the contribution of capillary pressure to the measured strength gain.• The relaxation behaviour of concrete indicates that the younger concrete is able to resist a larger amount of stress build-up compared to the older concrete.This therefore indicates that the relaxation behaviour of fresh concrete is dependent on the rate of hydration.• Multiple loading results indicate that the younger concrete is able to complete more loading cycles compared to the older concrete samples.The ability to complete multiple loading cycles gives an indication of the resilient nature and ductility of concrete during the stiffening phase.• The relaxation results also indicate that the mechanism behind the relaxation behaviour of fresh concrete is due to the negative capillary pressure build-up induced by the mechanically applied tensile strain.

Fig. 10 .
Fig. 10.Multiple loading results for test samples at 60 and 120 minutes.

Fig. 11
Fig.11displays the relaxation behaviour under multiple loadings for the 180 and 240 minute samples.The 180 minute concrete was nearing the end of the setting phase and the results show that the concrete was not able to relax all the induced mechanical stress.As a result, the remaining tensile stress after the initial relaxation during the first cycle was carried over to the next cycle, where the total load induced within the concrete sample is equivalent to 50% of the average maximum tensile strength, plus the remaining stress of the previous cycle.The 240 minute sample was only able to complete one cycle before failure.The main reason for this occurrence is believed to be due to the brittle nature of concrete, which lacks the ability to relax stress build-up compared to the 60 and 120 samples.The small amount of stress relaxed during the first cycle, was not sufficient in allowing for a second successful loading cycle.

Fig. 13
Fig.13shows a system of particles surrounded by water in three stages.Once the displacement of the actuator ceased (unloading), the high tensile force experienced during points of loading, starts to reduce as the particles rearrange in a way that decrease the tensile force measured in the gauge area.