1.1 Marine environment

According to NBR 6118: 2014 the marine environment is recognized as aggressive to concrete structures, where it is considered to be in aggressiveness class IV. Its influence on the durability of concrete structures depends on the microclimate in which it is found (^{Cascudo et al., 2014}). The aggressive agents present in the marine environment cause both the corrosion of the steel reinforcement and the corrosion of the concrete (cement matrix) (Lima and Morelli, 2004).

According to ^{Andrade (2001)} and ^{Medeiros (2012)}, the penetration of chlorides into the concrete can happen in different ways, one of which is the incorporation of the chloride into the mass of the concrete, which has become less likely due to the limitations imposed by the current standards. Examples of other mechanisms that occur most frequently are capillary absorption and diffusion, which depend on external factors.

According to ^{Verás Ribeiro (2014)}, capillary absorption is a mechanism where the chlorides present in a liquid medium penetrate into the concrete from the flow of the liquid, by the effect of surface tension acting on the capillary pores. According to ^{Verás Ribeiro (2014)} and ^{Meira (2009)}, this mechanism depends on the pore diameter, the surface tension of the liquid, its density, and its viscosity.

Diffusion occurs because of chloride concentration gradients. Chlorides in higher concentration regions move to regions of lower concentration (^{Meira, 2009}).

In the 2014 NBR 6118, attention can be drawn to a chapter on this subject, where classes of environmental aggressiveness for concrete constructions in urban and rural areas are identified as weak, moderate, strong, or very strong.

Such regulation also details the type of concrete to use as well as the specifications for the nominal covering to be adopted according to the component or element of reinforced or prestressed concrete. NBR 6118:2014 already recommends that the covering should be respected for construction projects including those with little or no control, since this is equal to the minimum covering necessary plus a tolerance factor.

Most of the constructions in direct contact with the marine environment begin to show pathological manifestations within a short time after construction, with significant decomposition of the concrete and intense corrosion of reinforcement (^{Verás Ribeiro, 2014}). Sea water has high amounts of chloride ions. The classification between aggressive and non-aggressive environments is relative and refers mainly to the amount of H₂S, SO₂, NOx, SO₄, and Cl^- present (^{Helene, 1986}).

For ^{Medeiros (2014)}, these oxides are extremely aggressive and contribute to the acceleration of the process of corrosion of the embedded reinforcements, even in small proportions. As a reference, it is possible for the corrosion rate in a marine atmosphere to be on the order of 30 to 40 times higher than that of a pure rural atmosphere (^{Verás Ribeiro, (2014}).

In cities that are subject to such saline exposure, constructions of concrete or reinforced concrete, whether in contact with sea water or not, always suffer pathological manifestations and mechanical degradation that require permanent maintenance (^{Cascudo et al., 2014}).

Some projects locate concrete structures directly in the marine environment: submerged or partially-submerged in sea water, within the high or low tide zone, the atmosphere zone, the splash zone, or the mist zone (^{Cascudo et al., 2014}).

Sea water is one of the elements of nature most damaging to concrete structures. In its composition, it is possible to find constituent elements of many chemical compounds.

According to ^{Lima and Morelli (2005)}, oceanographers have identified the elements distributed in ocean waters, as well as their various states and components of chemical compounds. Some compounds are stable, such as those containing sodium and potassium; while others are relatively unstable, such as those containing silicon and magnesium. The magnitude of concentration can be divided into three groups:

The composition of sea water, which can vary according to temperature, latitude, depth, and proximity to land, has been researched over a long period of time. Dittmar, in 1870, after analyzing thousands of water samples from all seas, found that, despite varying amounts of salts dissolved, the proportions of the main elements were constant. Salinity varies from 3.3% to 3.7% in the open ocean; with the overall average salinity for all oceans being 3.5%.

1.2 The City of Recife and Contact with the Sea

Recife, the capital of the state of Pernambuco, is popularly known as the “Brazilian Venice” because of its buildings along rivers and by the sea, and its numerous bridges that dominate the urban landscape, being one of the cities with most urban equipment in contact with sea water. Fig. 1 shows a panoramic view of the municipality of Recife.

Figure 1 a) Panoramic view of Recife, PE. Source: ^{Google Maps (2017)} Available at: <http://embrasil.s3.amazonaws.com/upload/ciudad/81C-37.jpg >. Accessed on: 11/07/2017. b) Map of Brazil with location of Recife, Pernambuco highlighted. 

Recife, because it is a coastal city, with a warm climate, high humidity, and predominance of the winds coming from the direction of the Atlantic Ocean, Recife’s concrete structures suffer aggressions from agents of diverse types of pathological manifestations. According to NBR 6118: 2014, Lima and Morelli (2004), ^{Verás Ribeiro (2014)} and ^{Medeiros (2014)} classified and defined four zones of aggressiveness that fit well with what happens to the existing concrete structures in the city. These regions are:

1.3 Mohr method

The need to identify how the chlorides penetrate into reinforced concrete structures is necessary because of the subsequent corrosion that it provokes. Several methods have been developed for this, including the colorimetric silver nitrate spray method, which is a qualitative test to identify free chlorides in concrete, and the Mohr method, which is a laboratory test using titration, also with silver nitrate.

According to ^{Mota (2011)}, the colorimetric silver nitrate spray method uses a methodology based on the application of a chemical indicator capable of altering the coloration of the concrete in the presence of chlorides. According to studies by Otsuki, Andrade, and Meck, this can vary based on factors such as cement type, water/cement ratio, and type of material used, as shown in Table 1.

This method was developed in 1970 by Dr. Mário Collepardi in order to verify the presence or absence of chlorides in concrete samples and, therefore, to be able to determine the penetration front of chlorides into the structures exposed to marine environments. This technique also contributed to the development of the process of fixation of free chlorides in the cement matrix (^{MOTA, 2011}).

To carry out the MOHR method, the 5% K₂CrO₄ (potassium chromate) solution, which acts as an indicator, is used to stain samples yellow, and then a solution of 0.0141 M silver nitrate (AgNO₃) is used until the liquid sample obtains a “brick red” color in titration combining with the free ions in the concrete sample.

1.4 Standardization: ISO TC71/SC 1(26-07-2010) and ISO/WD 1920-11

According to this ISO, reinforced concrete structures exposed to chlorides, whether from marine waters or other sources, must meet the durability criteria for which they have been designed for the entire life of the project. The possibility of corrosion of the reinforcement increases significantly as the chloride content gradually accumulates inside these structures. For this reason, the degree of diffusion or penetrability of the concrete represents important properties to be evaluated, whereas these technical specifications establish a test method that can be applied to samples prepared for the evaluation of the potential properties of resistance to chloride that a given concrete mix will present.

For this reason, the degree of diffusion or penetrability of the concrete represent critically important properties to be evaluated. These technical specifications establish a test method that can be applied to prepared samples in order to evaluate the potential resistance to chloride that a given concrete mix will present.

These technical specifications represent a method for determining the unidirectional chloride penetration parameters in the non-continuous state in preconditioned hardened concrete specimens.

2.1 Characterization of materials utilized

2.2 Molding the test specimens

Nine cylindrical concrete specimens (10x20 cm) were molded for three different mechanical strengths> low, medium, and high; with three specimens for each (See Table 2). After being removed from the mold, all specimens were submitted to wet curing for 14 days. One of each type remained in the wet curing until 28 days in order to maintain all types equal and to measure mechanical resistance.

The other six test specimens were divided in half, with their ends aligned and left in a dry environment for another 14 days, resulting in a total of 12 10x10 cm specimens. At 21 days they were waterproofed on all sides except for the top and bottom, as shown in Figure 2. At 28 days of age, they were submerged in fresh water, remaining there until 35 days, when they were removed and placed partially submerged in sea water with one of their ends on a fine bed of crushed stone for another 7 days, according to Figure 3. At 42 days, they were removed from the sea water, allowed to dry for 24 hours in the natural environment, and each identified by type of concrete and the division made.

Table 2 shows the unit ratio of the materials used in relation to the cement, by mass. The slump values were kept fixed at 160 ± 20 (mm), by varying the amount of dry materials.

Figure 2 Cutting, naming, and painting the test specimens. 

Figure 3 Seawater test (schematic design). 

After completing these processes, holes were drilled in each test piece on the end that was in contact with sea water and were drilled from largest to smallest with depths of 1, 2, and 3 centimeters, respectively. An impact drill was used to make the holes, a ruler for measuring, a bench drill for greater precision and drill bits for concrete of gauges 14, 10, and 8 mm. From each hole, all material (powder) was collected and packaged in individual bags, named and numbered for identification when tested in the chemical laboratory (see Figure 4).

Figure 4 Drilling tests with fixed drill. 

2.3 Performing the Mohr method

The second part of the study was carried out in the chemical laboratory at the Catholic University of Pernambuco, under the guidance of Professor Sérgio Paiva. The purpose of the tests was to use the Mohr method, titration by silver nitrate, to determine the amount of chloride ions that penetrated to the depths drilled in the specimens (1, 2, and 3 cm).

In Table 3, it can be seen that 45 samples were cataloged, 36 of the partially immersed specimens and 9 of the non-immersed specimens.

Parts of 45 samples were placed in beakers, weighing 0.5 grams. After this, the drilled powder from each test specimen, parts A and B (two pieces of 10 x 10 cm), was placed in three flasks and numbered from 1 to 18, with a pipette of 50 ml distilled water, and stirred to mix.

Then this heavy "powder" was diluted with distilled water in a 250 ml volumetric flask. The contents of the volumetric flasks were left to rest for at least 24 hours and were then diluted with distilled water to 250 ml and shaken individually to homogenize them.

Of this 250 ml, three samples of 10 ml each were withdrawn and placed in flasks. In each of these samples, 1 ml of potassium chromate (K₂CrO₄, 5%, 99% pure) was placed, which worked as indicator and left a yellow coloration. After this, each sample was titrated with silver nitrate (AgNO₃ - 0.0141 mol/l, 97.8% pure) until reaching a red brick color. This would show the amount of silver nitrate needed to associate with the free chloride ions in the cement mass of each sample.

The results were plotted individually in spreadsheets and the amount of chloride ions in each sample that reacted with the cement mass was calculated according to the ratio (poor, medium and rich) and depth of penetration (1, 2, and 3 centimeters). The process is detailed in Fig. 5.

Figure 5 Diagram of the tests carried out in the chemistry laboratory