Durability Indicators in High Absorption Recycled Aggregate Concrete

The use of recycled aggregates in structural concrete production has the inconvenience of increasing the fluid transport properties, such as porosity, sorptivity, and permeability, which reduces the resistance against penetration of environmental loads such as carbon dioxide and chloride ion. In this paper, behavior of ten concrete mixtures with different percentages of coarse aggregate replacement was studied. The recycled material was recovered by crushing of concrete rubble and had high absorption values. The results showed that it is possible to achieve good resistance to carbonation and chloride penetration with up to 50% replacement of recycled coarse aggregate for 0.5 water/cement ratio. Finally, new indexes for porosity and sorptivity were proposed to assess the quality of concrete.


Introduction
Concrete is a fully recyclable material, provided it is not contaminated.Recycled concrete has been used in many nonstructural forms, even for treatment of acidic groundwater [1].It can be used as structural material as long as it complies with the specifications for normal concrete.However, the manufacture of concrete elements with recycled aggregate from construction and demolition debris can have the drawback of high porosity and permeability, which facilitates the penetration of environmental loads such as chloride ions and carbon dioxide, causing chemical reactions that lead to corrosion of reinforcing steel, thus affecting its durability.
The durability of concrete is defined as the ability to withstand the effects of time, chemical attack, abrasion, or other degradation process.Durable concrete must retain its original shape, quality, and service conditions when exposed to the environment [2].
A variety of tests are available for the study of the durability of concrete, which require specialized equipment and methods of study, as well as long periods of exposure to observe deleterious effects.In the case of recycled aggregate concrete, several researches have been focused on the study of mechanical properties, often omitting durability aspects.
In this work, three fluid transport properties were determined in concrete indirectly related to durability, such as porosity, sorptivity, and air permeability.Two direct indicators were also analyzed, carbonation, and chloride penetration.Results are compared with indexes reported in the literature to determine the amount of coarse aggregate that may be replaced.

Porosity
The deterioration of concrete by non-mechanical causes is mainly due to the presence of water with dissolved harmful substances.Penetration and transport of water into the concrete takes place through the pores and cracks in the material and depends on the type, size and distribution of such voids.When using recycled aggregates in concrete 2 Advances in Materials Science and Engineering Porosity has a great influence on the mechanical and durability properties of concrete.It represents the total voids in the material (cement paste plus aggregates).Porosity of the cement paste is the sum of the capillary and gel pore volume and represents the voids from solid components of the hydrated cement paste.In aggregates, the pore size varies in a wide range, but even the smallest pores are greater than gel pores in cement paste.Some pores are totally immersed within the particle, but others are open to the surface allowing the entry of water and other aggressive agents [4].Given that the aggregate represents three quarters of the concrete volume, the rock's quality would be an important factor in porosity.However, it is the capillary porosity the one that influences the concrete durability as the cement paste surrounds the aggregate particles.According to CEB [5], the quality of the concrete in terms of porosity is classified as good if it is below 15%, moderate if it is between 15 and 17%, and poor if it is over 17%.

Sorptivity
Sorptivity in a concrete sample gives a value of the effective porosity.It represents the increase in mass of the specimen as a result of water absorption by the capillary pores.It can be an important parameter to measure performance of concrete to the effect of environmental loads.Ho and Lewis [6] recommend a value less than or equal to 3 mm/h 0.5 for up to 30 mm coatings in severe conditions and up to 6 mm/h 0.5 in less aggressive environments.Other authors such as Olorunsogo and Padayachee [7] have rated sorptivity for recycled aggregate concrete in ranges shown in Table 1.

Permeability
Permeability tests on concrete are made to determine a coefficient describing the flow of liquids or gases through the pores of the material.A method available is the air permeability, included in the Swiss Standard SIA 262/1-GB: 2003 [8].It is based on creating a vacuum within a double chamber cell which is placed on the concrete surface and measuring the speed with which the pressure returns to the atmospheric value.The application of this method allows to obtain the air permeability coefficient () describing the quality of the concrete surface according to the ranges given in Table 2.

Carbonation
In concrete structures, a major cause of deterioration is the reaction between the hydrated products of the cement paste and the carbon dioxide present in the atmosphere.This reaction reduces the pH of the pore solution, thus resulting in corrosion of the reinforcing steel bars which in the end decreases the service life of the structure.The resistance to a reduction in pore solution pH depends on the cementitious binder and it is strongly related to its acid neutralization capacity [9].Since carbonation process can be very slow, for research purposes, the use of an acceleration chamber with controlled internal environment was chosen [10].Carbonation depth are usually determined by colorimetry spraying an acid-base indicator on a freshly cut sample.According to literature [11] carbonation coefficient must not exceed the critical value of 6 mm/año 0.5 ; coefficients above this value denotes low carbonation resistance.

Chloride Penetration
Chloride ion (Cl − ) is the principal agent capable of causing corrosion process on steel reinforcement embedded in concrete elements and thereby reducing its durability.Chloride penetration in concrete can occur by diffusion if the element is submerged in salt water or by capillary suction if it is exposed to marine aerosol.In laboratory, the effect can be induced by immersion of the samples in synthetic sea water, by exposure in a salt spray chamber or by migration of chlorides, forcing its entry under the action of an external electric field.For durability purposes, the accepted Cl − concentration should not exceed 0.4% by weight of cement [11].

Materials.
The experiment consisted of a study of ten concrete mixtures combining two water/cement ratios (w/c) with five replacement rates of recycled coarse aggregate (%).The w/c ratios were chosen to represent the highest w/c suitable for durability (0.5) and the highest w/c suitable for strength (0.7).The materials used were OPC, water, coarse aggregate and fine aggregate, obtained by crushing limestone from a local quarry, and recycled coarse aggregate recovered by cutting and grinding heterogeneous structural concrete rubble.Aggregate properties were determined according to ASTM Standards [12].They can be seen in Table 3. Highlights are the low specific gravity of all aggregates, as well as the high absorption values (above 6%).However, according to Solís-Carcaño and Moreno [13], these properties for normal and recycled aggregates can be located in the boundary between limestone materials of good and poor quality.

Mixture Design.
Mixture design was performed based on the ACI method [14].Each concrete mixture was designated with NC (normal concrete) or RC (recycled concrete) followed by % (25, 50, 75, or 100).The last digit was added to define w/c (5 for 0.5 and 7 for 0.7).Effective absorption of 80% for the mixing water correction was considered.The relative amounts of the materials for each mixture, before daily moisture corrections, are indicated in Table 4.

Test Specimens.
The concrete batches were 120 liters made in 8 minutes.Cylindrical specimens of 10 × 20 cm were cast for porosity and sorptivity tests and 7.5 × 15 cm for carbonation tests.Also, prismatic beams of 10 × 15 × 60 cm were cast for testing air permeability and penetration of chlorides.The specimens were subjected to a process of moist curing by immersion for 28 days and then conditioned at lab environment (80% RH and 28 ∘ C) for 90 days before carrying out tests for porosity, sorptivity, and air permeability.
7.4.Porosity, Sorptivity, and Air Permeability Tests.Porosity was determined using dry weight, saturated weight, and submerged weight of five test specimens following ASTM C 642-06 [15].For each sorptivity test 3 slices, 5 cm thick, were cut.In each slice a layer of epoxy paint was applied on all sides except for the exposed face to allow the entry of water by capillary suction in one direction.Norwegian method purposed by Fagerlund [16] was used.The air permeability tests were performed with double vacuum chamber equipment on three beams for each mixture and in three areas for each exposed face.

Carbonation and Chloride Penetration
Tests.The specimens were exposed to carbonation for ten weeks in an acceleration chamber with 65% relative humidity and 4% CO 2 .The specimens were removed progressively from the chamber after 14, 35, 56, and 70 days of exposure and cut with a chisel and hammer.Carbonation depth measurements were performed using a digital caliper after spraying an acid-base indicator (1% phenolphthalein solution) according to RILEM procedure [17].
The specimens for chloride penetration were immersed for ten weeks in synthetic seawater, prepared in a similar manner as specified in ASTM D 1141-98 [18].Except for the exposed face, all specimens were precoated with epoxy paint, to allow entry of chloride in only one way.After that period the specimens were drilled on the exposed face in eight layers, to extract concrete dust at different depths up to 6 cm.The obtained powder was subjected to digestion with nitric acid for the total chloride extraction.The amount of chloride by weight of cement was determined following the FDOT method [19].

Results and Discussion
Results of porosity and sorptivity for each mixture according to % are shown in Figures 1 and 2. All porosity values were very similar and above the limit of 17%, which indicates a poor quality concrete.These results are higher than the 12.5% reported by Buyle-Bodin and Hadjieva-Zaharieva [20] but similar to the 18-20% reported by Gómez-Soberón [21] for 0.5 w/c.However, in the case of aggregates with high absorption, porosity is not a conclusive indicator of the concrete quality, and the use of other direct and indirect evidence is necessary to determine its performance in terms of durability [22].
Sorptivity was calculated with the following [16]: where  is the sorptivity (mm/h 0.5 );  is the resistance to water penetration (s/m 2 );  is the required time for capillary  The sorptivity values increased slightly when % increased.NC5 control concrete was the only below 10 mm/h 0.5 .The remaining results for 0.5 w/c and those with 0 and 25% for 0.7 w/c were in the range of 10-15 mm/h 0.5 , classified as poor quality.The others were rated as very poor, above this range.Other researchers [7] have found similar results for equivalent conditions (15-20 mm/h 0.5 ).
In Figures 3 and 4 the air permeability coefficients () and carbonation coefficients converted to atmospheric values (  ) for the different % are shown.In both tests the difference between the values from concrete specimens with 0.5 w/c to those with 0.7 w/c is noteworthy.
The model developed for the air permeability coefficient is based on the Hagen-Poiseuille law for elastic fluids, as shown in the following [23]:  where  is the coefficient of air permeability (m 2 × 10 −16 );   , the volume of the inner cell system (m 3 ); , the crosssectional area of the inner cell (m 2 ); , the viscosity of air at 20 ∘ C (2 × 10 −5 N s/m 2 ); , the estimated porosity of the cover concrete;   , the atmospheric pressure (N/m 2 ); Δ  , the pressure rise in the inner cell at the end of the test (N/m 2 );   , the time at the end of the test (s); and  0 , the time at the beginning of the test (60 s).
Since the used equipment in these tests was calibrated by default to constant values of porosity equal to 0.15, it was necessary to adjust the results provided by the device using actual porosity values that were previously determined.
The air permeability coefficients increased from 24 to 64% as the % increased.0.5 a/c concretes with 0, 25, and 50% were in the range of 0.1 <  < 1.0 m 2 × 10 −16 , that is considered normal quality.All other mixtures had values above 1.0, indicating poor quality of the concrete surface.
Carbonation coefficients (  ) were obtained following the simplified model shown in the following [10]: where   is the carbonation coefficient converted to atmospheric value (mm/y 0.5 );   is the average of accelerated carbonation coefficients (mm/y 0.5 );   is the CO 2 concentration outside the chamber (0.051%);   is the CO 2 concentration in the chamber (4%);   is the coefficient for the progress of carbonation depth (mm/y 0.5 );   is the carbonation depth (mm);  0 is the initial carbonation depth (mm); and  is thenumber of tests performed during the exposure period.
values increased according to the increase of %.0.5 w/c concretes with 0 to 75% achieved acceptable values between 4 and 6 mm/y 0.5 ; for the other mixtures,   was above that range which shows poor resistance to carbonation.
In Figures 5 and 6 the progress of carbonation depth versus time is modeled.For 0.5 w/c concretes, the worst result was for RC100-5, where at 25 years, carbonation depth would be 32.5 mm.For 0.7 w/c concretes, carbonation depth at the same age would reach much higher values.These figures highlight the similarity between the carbonation behavior of mixtures RC100-5 and CN7; thus the worst 0.5 w/c mixture equals the best 0.7 w/c mixture.
In Figures 7 and 8, the relationship between the   values with  and  is analyzed, respectively.The correlation coefficients () were greater than 0.9 indicating a relatively strong dependence between the studied variables in both cases.
where %Cl is the chloride concentration by weight of concrete,  is the intersection point of the potential's regression line with the -axis of a Gran plot (mV),  is the silver nitrate normality, and  is the sample weight (g).The results were multiplied by the concrete/cement ratio to obtain the Cl − concentration by weight of cement.Mixtures with 0, 25, and 50% reached chloride concentrations below permissible limit of 0.4% for 0.5 w/c.Mixtures with 75 and 100% and all the 0.7 w/c mixtures were well above the limit of 0.4%.In all cases, a change in slope was observed at 2.5 cm depth, where there is a significant increase in the chlorides concentration.
The diffusion coefficients were calculated with (5), which gives solution to Fick's 2nd law in nonsteady state.Consider   where  is the chloride diffusion coefficient (cm 2 /s);  is the chloride surface concentration;  is the initial chloride concentration in concrete;  is the chloride concentration at depth ; and  is the exposure time.
The results for each mixture can be compared in Figure 11, which shows that the diffusion coefficients were increasing according to %.The difference between 0 and 100% was very strong (89%) for 0.5 w/c, while in concretes with 0.7 w/c it was only 10%, which is attributed to poor quality concrete where % do not have significance for these mixtures.Otsuki et al. [27] had reported differences of 17% for 0.55 w/c and 5% for 0.7 w/c.

Figure 5 :Figure 6 :
Figure 5: Calculated progress of carbonation depth versus time for 0.5 w/c.

Table 1 :
Suggested ranges for sorptivity in recycled aggregate concrete.

Table 5 :
New suggested indexes for porosity and sorptivity in recycled aggregate concrete.