3.1. Methodology

The present case study was developed based on the pathological problem-solving methodologies proposed by (^{Lichtenstein, 1985}) and the facade ceramic cladding diagnosis and recovery proposed by (^{Campante, 2001}).

Four inspections were carried out at the site in order to observe the existing pathological manifestations, from which were also sought the supporting evidence and case history to understand the causes and evolution of the problems over time.

Additional adherence strength and percussion tests were carried out in order to support the diagnosis. Following this, the diagnosis was made based on all the collected information, as well as the prognosis, which associates each symptom with its causes and possible methods of recuperations.

It should be noted that the results of the tests were obtained from a specialized company who was contracted by the building owners, as a result of the detachment of some tiles from the facade cladding.

3.2. Building inspection: collection of supporting evidence and case history

3.2.2 Description of facade and cladding

The rough mortar utilized for the facade was composed of cement and coarse sand. The plaster mortar was composed of cement, gravel and coarse sand, with the presence of clay demonstrated by the reddish color of the mortar, a common practice at the time of construction.

The ceramic tiles used were of the semi-stoneware type (water absorption between 3% and 6%, medium) and medium thickness of 6 mm, with dimensions of 5x25 cm. The striped white marble tiles (water absorption between 0.1 and 0.4%, medium to high) had a height of 95 cm and width ranging from 20 to 30 cm. The grouting between the ceramic tiles was done with mortar composed of cement and fine sand, with a width of 15 mm, and that for the marble with white cement approximately 2 to 3 mm wide.

As shown in Figure 1, the facade was divided into eight sample areas (section T1: east facade; sections T2, T3, and T4: south facade; sections T5, T6, and T7: north facade; and section T8: west facade).

Figure 1 Sketch of building section identifying the facade and its division into sample areas. 

It should be noted that the city of Recife has a humid tropical climate of monsoons with high relative humidity and high rainfall. On the hottest summer days, the temperature can reach 35ºC, while on colder days the minimum temperature is 16ºC. The average annual temperature is 26ºC, probably more common during the installation of the facade cladding. Recife has approximately 2,550 hours of sunshine per year, with an ultraviolet index (UI) of solar radiation at the Earth's surface of 10 on a scale of 14, considered to be very high.

Figures 2, 3, 4, and 5 illustrate the cladding and architectural details of the facades.

Figure 2 Northern facade (sections T5, T6, and T7). 

Figure 3 Eastern facade (section T1). 

Figure 4 Southern facade (sections T2, T3, and T4). 

Figure 5 Western facade (section T8). 

The north facade (Figure 2) contains details designed to contribute to interior ventilation, where airflow enters indirectly below the aluminum frames, controlled by a "guillotine" mechanism attached to the marble windowsill (Figure 6).

Figure 6 Schematic detail of indirect ventilation in window frames. 

This detail is important in the analysis of facade pathologies, since most of the occurrences of detachment of the marble slabs were found in this area at the bottom of the beam.

3.2.3 Problems identified

During the inspections, a high incidence of detachment of marble slabs was noted on the north facade, where the structure suffered greater exposure to aggressive environmental agents (Figure 2).

In general, the marble slabs and building facade joints showed signs of moisture condensation, highlighted by the presence of molds. The front facade of the building faces towards the sea (east), thus having its cladding exposed to continuous direct sea breezes. This cladding is entirely composed of striped white marble and has spots of black coloration at variousl points, characterized by the formation of mold caused by the constant moisture in the grout and the marble slabs due to the water absorption and porosity (grout), as well as the favorable environment for the proliferation of fungi (Figure 3).

The southern facade is composed of semi-stoneware ceramic tiles of beige color and rectangular plates of white striped marble, with dimensions of 95x20 cm and 95x30 cm (Figure 4). On the south facade, grout and plates with black spots were observed at roof level, characteristic of mold. There is also a history of detachments occurring in the cladding applied horizontally from below the bottom of the edge beam of the roof balcony.

The west facade is covered mostly by the same ceramic as the other facades, in a closed sheet, having only at the edges a single finishing row of white striped marble. On this facade, the detachment of two slabs can be seen in the finishing marble, while the ceramic remains intact without showing signs of adherence failures (Figure 5).

Another occurrence reported with respect to the west facade was the recurring appearance of infiltrations on the 8th floor, according to a resident. In addition, the condominium contracted for 15 years a company to perform maintenance services on the facade, recovering the grouting of each facade and replacing some tiles that showed signs of adherence failure.

The technique used to lay ceramic and marble tiles was characterized by the fact that it was applied simultaneously with the application of plaster, commonly known as Brazil as the bolão method (Figure 7).

Figure 7 Example of plaster mortar spreading and fixing on the back of the tile - bolão method 

The cement paste was spread on the back of the tile. In this type of application, the previous wetting of the cladding material (ceramic tile and natural rock - marble) is an important condition to avoid the loss of part of the water needed to cure the mortar.

No joints were observed that would act as movement joints. Most of the occurrences of detachment happened with tiles set in the horizontal, but there was also detachment of some marble slabs set in the vertical panels, confirming the existence of failures in marble slab adherence of the facade cladding. After the removal of a piece of concrete from the bottom of the periphery beam, the corrosion of the reinforcement and the separation of the covering from the reinforcement became evident.

On the south facade, specifically on the first floor balcony, after the removal of some cladding stone, it was observed that the reinforcements and supports of the beam are in the process of corroding. These issues will not be the subject of discussion of this study.

3.3. Complementary tests of adherence strength resistance and percussion

Six direct adhesive strength tests were carried out in each sample area of the facade, for a total of 42, distributed across different floors. For confirmation of each point to be tested, it was important to have the percussion test performed by a professional with more than 20 years of experience at the specialized contracted company - a laboratory accredited by the National Institute of Metrology, Quality and Technology (INMETRO), homologated by the Ministry of Cities as a Technical Evaluation Institution (ITA), and capable of verifying the performance of innovative construction systems or those not yet disseminated widely in the Brazilian market.

A criterion adopted for carrying out the strength tests was to replace the sample when the hollow sound was detected, so as not to invalidate the experiment, because the low result could compromise the data obtained.

The tests were performed according to Brazilian standard NBR 13755 (^{ABNT, 1996}), with the cut in the cladding to prepare the sample done by marble saw, which cut through the cladding layers, making it possible to identify the most fragile interface of the rupture. The percussion test was carried out in order to identify areas lacking adequate adhesion between the mortar and the tile.

3.4. Presentation and analysis of results

The test results can be verified in Figures 8 and 9. According to Brazilian norm NBR 13755 (^{ABNT, 1996}), for every six points verified in the same section, at least four of them must have a rupture tension (traction adherence) value points greater than 0.30MPa.

Figure 8 Results of the tensile adherence strength tests (^{Tecomat, 2011}). 

Figure 9 Results of the percussion tests (^{Tecomat, 2011}). 

It should be noted, initially, that the sections tested (T) were clad with marble slabs, except for section 3, which used ceramic tiles and whose results were satisfactory. Judging by the standardized criteria, the values obtained show clear defects in adherence resistance in sections 2, 5, 6, and 7.

In the case of section 1, contrary to expectations, results were found to be far better than the criteria mentioned above. It is reasonable to suppose that what could justify such a large difference in performance may have been either the use of different procedures for this section, or greater ability on the part of the professional who executed the east facade.

Percussion tests were also performed, where a hollow sound obtained by means of striking the cladding with a rubber hammer indicates a flaw and suggests the possibility of detachment of the tile. It should be clarified that there is no normative criterion that indicates acceptable levels for this test and it should not be adopted in isolation (Table 1).

On the side of the first-floor balcony on the south facade, discrete rupture points at the base/slab interface can be seen, which is likely to be caused by inadequate cleaning of the base before the cladding was installed.

The reddish color of the plaster mortar indicates the presence of gravel in the mortar, a material widely used at the time (to substitute hydrated lime and ensure plasticity with lower cost), which today is avoided due to the difficulty in controlling the quality of the material and thus presenting a high degree of variation in its composition. The presence of the clay in the mortar may be responsible for the detachment or poor adhesion, due to very fine grains of clay penetrating the pores of the substrate in place of the cement hydration products that would form at the interface and produce the mechanical locking of the paste in the pores.

Analyzing the execution process used at the time, where quality depended very much on the skill of the professional, in the case of the adherence of stone cladding, the risk of failure occurring was quite high. At the time, there were no premixed or additive products, and regularization mortar (plaster) was used, to help fix the cladding elements, especially in the case of large tiles or slabs.

In order to start the fixing process, it was necessary to set up wooden shoring to support the first layer of tiles, which would later serve as support for the subsequent layer and so on. Due to their weight, the tiles were supported by the lower layer, while application of the mortar occurred simultaneously with the percussion of the plates using a rubber hammer, until the correct position was obtained. To facilitate this activity, the mortar was not spread across the entirety of the back of the tile, allowing it to be spread and ensure complete filling during the percussion, often unsuccessfully.

It is also necessary to consider the possibility of a gradual reduction in adherence between the back of the tiles and the mortar, due to eventual natural settling of the building (wind action, temperature and, in particular, concrete inflow and slow deformation of the structure through time), which may also contribute to this problem.

In general, the results of the direct tensile strength and percussion tests indicate that most of the marble cladding area is found to have compromised adhesion in some stretches with imminent risk of detachment. The perspective is that with movement, whether due to rain, wind, or the various stresses to which they are subjected, other cladding tiles will eventually detach, producing a hazardous situation with the potential for material and personal damages.

The black stains on the marble cladding in the wetter areas have the appearance of fungi (mold and mildew) that develop in the pores of the cladding and grout, creating a crust on the surface. Regarding the procedures to be adopted to recuperate these pathologies, different solutions must be adopted for each case diagnosed. However, it is important to remember that they are interconnected, as there is no way to treat the facade cladding without performing structural recovery of the compromised pieces.

Specifically, stains caused by atmospheric contamination (pollution) and humidity should be treated after recovery of the structure, cladding, and grout, by cleaning with pressurized water.

The elaboration of a cladding project is recommended to replace the mostly compromised facade marble by other material to be specified in order to reconstruct facade and provide greater efficiency in its functions.

For preventive maintenance purposes, an annual inspection is recommended in order to monitor any new occurrences.