2.1.1. Steel frames

The cold shaped metal frames (tensed C shape) have height of 90mm and width of 40mm with thickness of 12mm, produced with galvanized steel. The steel has a yield stress of 600MPa. The steel frames received a surface treatment of zinc and aluminum, which ensures a high corrosion resistance.

2.1.2. Wood particleboars

The wood particleboards are structural panels of oriented strand board (OSB) wood type comprising thin strips and oriented perpendicularly to each other in several layers, which increases its rigidity and strength. These strips are attached with a resin applied under high temperature and pressure, phenolic type. The boards have thicknesses of 11.1mm and density 650.00kg/m³.

2.1.3. Cement boards

The cement boards are used to protect the wall system (external area of the building) and are made from a mixture based on cement and synthetic fibers. The boards have thicknesses of 10.0mm and density of 1.20g/cm³.

2.1.4. Plasterboards

The plasterboards have a density of 1.45g/cm², are non-combustible and have flexural strength of 9MPa. Two types of plasterboards were used. The board named Standard, which is ideal to use only in dry environments is made of natural gypsum. The other board, named "fire resistant", is produced from natural gypsum with the addition of natural fibers and admixtures to improve its resistance to fire. 2.1.5. Glass wool

The glass wool is used for a better thermal and acoustic insulation and has a thermal resistance of 1.52m².°C/W, density 40.00kg/m³ and a thickness of 50.0mm.

The samples have 315cm of width, 300cm of height and 13.5cm of thickness. The steel frames used in the assembly of this system were spaced every 60cm, with a thickness of 90.0mm. The arrangement of the frames was chosen according to the ultimate bending moment resistance of the internal wall system (plaster and wood boards), assuming a safety factor of 1.4.

The sample labeled "A" is a wall with OSB composite board, 11.1mm thick, applied both internally and externally. This board has a bracing structural function of the boards. The plasterboard used was the standard type, with 12.5mm thick, applied over the OSB board. In the external surface, also on the OSB board, a cement board with thickness of 10mm was installed. Sample "B" was composed of the same materials than sample "A", except the plasterboard, which was the fire resistant type, 12.5mm thick.

Table 1 gives a breakdown of the constituent layers of these samples. Fig. 2 shows the steps of mounting and the appearance of the finished sample. Fig. 2a stands out the installation of the steel frames (rigid frame), Fig. 2b details OSB boards application on halfway up to the plasterboard, and Fig. 2c is the finished sample coupled to the vertical furnace.

The vertical furnace is heated by four liquefied petroleum gas burners, controlled by differential pressure as the examined area. The burners are installed at the furnace sidewalls (two per wall) and are calibrated to increase the temperature according to the heating curve established by [1]

The temperature development is monitored during the test by two thermocouples installed inside the furnace. In the event of a deviation, a correction of the flame intensity produced is performed automatically. The temperature of sample’s external surface is measured by five thermocouples, also installed within the furnace. The temperature of sample’s exposed surface is the average of the temperature measured by thermocouples arranged symmetrically on the furnace inner region, in number equal to five, with one for each 1.5m² of exposed surface. Temperatures are measured with an accuracy of ±1.5%.

Table 1

Source: The authors.

Figure 2 Source: The authors.

Table 2

Source: The authors.

Figure 3 Source: The authors.

Figure 4 Source: The authors.

Table 2 presents the equipment used during the tests.

The internal (face of the sample exposed to fire) and external thermocouples (face of the sample not exposed to fire) are type K with diameters of 4 mm and 2 mm, respectively. Fig. 3 and Fig. 4 illustrate the positions of the thermocouples.

For the analysis of both systems, it was evaluated the parameters of integrity, thermal insulation and structural stability, as defined [14-17] to check the fire resistance of structural walls.

The analysis of the thermal insulation is made by the thermocouples measurements in the surface not exposed to fire of the sample. In order to be thermal insulated, the maximum rise of temperature must be less or equal to 180°C in each thermocouple or 140°C on the average of all the thermocouples, whichever occurs first, adding the initial temperature of the trials [14,16,17].

The cracks found in the samples during the test are not allowed to release hot gases through its openings. The integrity is verified with a dry cotton wool pad positioned on the crack for 10 seconds, which can not inflame during the exposure, as defines [14-16]. Cracks that have not caused the inflammation of cotton wool pad were disregarded in this study, as it does not compromise the integrity parameter. To assist the identification of the cracks during the tests, a thermographic camera was used on the surface not exposed to fire.

The structural stability was verified by the horizontal deformations developed during the test, which was measured using a laser tape pointed to the center of the not exposed surface. The system was subjected to the action of a uniformly distributed load of 15kN/m, simulating a slab action load. According [14], the fire resistance time of the sample corresponds to 80% of the time period in which the element was charged and subjected to the test. Regarding the overall deformation of the load-bearing wall system, the standard states that it can not be excessive, admitting a maximum deformation of 1/30 of the span.

The temperatures recorded in the sample "A" and Sample "B" are presented in Fig. 5 and Fig. 6, respectively.

Manifestations occurring during the test are presented Table 3.

Figure 5 Source: The authors.

Figure 6 Source: The authors.

The ambient temperature at the test beginning was 24.5°C and in turn the test limit is 164.5°C for the average temperature of five external thermocouple and 204.5°C for each thermocouple (thermal insulation).

The sample “A” test lasted 51.5 minutes, whereas the sample “B” lasted 62.0 minutes when the tests were finished. However, the average temperature of the five external thermocouple has not reached 164.5°C in both cases, or any individual thermocouple temperature reached 204.5°C. Therefore, the sample with fire resistant plasterboard had test duration 10.5 minutes, or 20.3%, higher than the sample with standard plasterboards.

According to the pictures and temperature graphs presented on Table 3, the samples had a similar behavior up to 40 minutes of test. The wall using the standard plasterboards (sample “A”) had initially a heat concentration at the joints, being later distributed throughout the system’s surface. The same occurred with the sample with fire resistant plasterboards (sample 'B'), which, however, presented a better performance for 10.5 minutes when compared to sample “A”.

In both cases, it was noted that the metal bolts are the first elements to present an abrupt heating. It was evident that its protection, or substitution by insulating material, can contribute to the thermal insulation of wall systems using plasterboards. The bolts and joints might be considered a weak point of this kind of system.

After 45 minutes from the beginning of the test, it was noticed a vertical crack between two cement boards (external) on sample “A” (Fig. 7). At the same moment, the cotton wool pad test was performed. With the crack propagation, the test was repeated at 51.5 minutes and, at this occasion, the hot gases expelled caused the cotton wool pad inflammation, characterizing the system’s loss of integrity.

In the test of sample “B”, vertical cracks were identified between boards (at the joints) 50.0 minutes after the beginning of the test (Fig. 8). However, trhough the cotton wool pad test the sample still maintains its integrity. After 62 minutes of testing, the action of the hot gases produced by this opening has passed to compromise system’s integrity and the test was finished.

Table 3

Source: The authors.

Note that the joints have proved to be the weakest point of the system, independently from the type of plasterboard used. For this reason, it is emphasized the importance of ensuring the fire resistance properties in all components of the system, not only at the plasterboards, as well as its joints.

Figure 7 Source: The authors.

Figure 8 Source: The authors.

Fig. 8 presents the evolution of measured deformations during both tests.

Since the exposed span length was 2500.0mm, the deformation limit is 83.3mm. For both samples, the deformation was very similar up to 30.0 minutes of test and less than 10.0mm. From the 30.0 minutes, sample “A” presented higher total deformation, leading to a loss structural stability. However, the deformations did not compromise the use of both samples, even though sample “A” reached 80.0mm at 50.0 minutes of testing.