Application of numerical modelling to evaluate the influence of smoke extraction flaps on fire safety of the buildings

The aim this paper is to present the simulation of the building in the computer program PyroSim by Thunderhead Engineering based on the FDS calculation engine. The article presents simulations of the use of smoke vents and their impact on fire safety in a public building based on the Fire Dynamics Simulator (FDS) calculation engine. Two scenarios were analysed in the simulation. The first simulation shows 180 seconds of fire without the smoke vent, while the second simulation shows 150 seconds of fire with the smoke vent open. A fire power density of 500 kW/m2 was adopted and is used in each case as well as the fire development coefficient, which was calculated on the basis of literature data. On their basis, simulations were carried out, the results of which were used to draw conclusions.


Introduction
In public buildings, due to the large number of people moving, special applications of fire protection are used.This is primarily due to the safety aspects of the people staying there and the protection of highvalue items.Technical and construction regulations are the basis for ensuring safety in such buildings.In order to meet the objective of adequate fire protection, public buildings are equipped with passive and active fire protection systems [1][2][3].One of the applications that are currently being implemented is the use of smoke vents, which are designed to remove smoke and heat during a fire [4,5].They are an extremely important element of any smoke exhaust system.The dampers connected to the control panels of the smoke exhaust system can be automatically opened to exhaust fumes and heat from the room, giving time for evacuation of people who are within the range of fire fumes.In the adopted simulation, two research analyses were performed.In the first, open smoke vents were used, while in the second, the simulation was carried out without open smoke vents.Using Thunderhead Engineering's PyroSim software to create the model allows you to understand the result, thanks to its readability.They are presented graphically.It is a very advanced scientific tool that can be used as a visualization of the calculated values before and during the modelling process [6][7][8][9].The calculations of the program show the values determined for each of the volumes into which the computational space is divided.The input data includes, for example, the building geometry made up of rectangular blocks, the properties of the materials present in the building, the measurement points to collect the results after the simulation and, above all, the combustion kinetics [10,11].

Models assumptions
This article presents a simulation of a fire in the library of the Faculty of Environmental Engineering building at Nadbystrzycka 40B street.Based on the data contained in the construction design, a 3D model of the room with dimensions of 7.41 m x 7.41 m x 2.75 m was created (figure 1).In order for the simulation to be carried out reliably, the fire power density was assumed depending on the type of room, using the table [11].In the presented analysis, the fire power density was assumed as for the library, i.e. 500kW/m 2 .It is a room with a strong character of fire development, the development factor is: 0.04660.The area of the burning elements is approximately 1 m 2 .Substituting all data into the formula (2), the time in which the fire reaches its maximum power will be 103.59s.Knowing the characteristics of fire development, the coefficient was determined using the table [11].Below is the formula that was used to calculate the fire development time.
Where: Q-power of fire T time to reach full fire power from the start of the fire γ-coefficient of fire development rate By rearranging this formula, let's calculate the time after which the fire stops growing: Ten exhaust vents with dimensions of 20 cm x 20 cm were installed on the walls in their upper part.A door with dimensions of 1.9 m x 2.25 m leads to the room.There are bookshelves in the room, two tables with computer stations and a desk for library employees with a computer and printer.

Materials and methods
The basic tool used in this article is PyroSim software by Thunderhead Engineering.Thanks to the software, a numerical simulation was carried out using the FDS calculation engine.Two fire development scenarios were adopted during the simulation.In the first scenario, it was assumed that the smoke vents remain closed throughout its duration.The fire power was assumed in accordance with the literature analysis, i.e. 500 kW, and the time of its development to reach the maximum power was calculated on the basis of formula (2) and is 103.59 s.The duration of the simulation was 180 seconds.
In the second scenario, a scenario was adopted in which the smoke vents are open for the entire duration of the fire in a public building, which is a library.The power of the fire and the time of its development were assumed to be the same as in the first scenario.The duration of the simulation has been reduced to 150 seconds.It was assumed that the source of the fire was three computers located in different locations in the simulated room.To obtain better results, it was assumed that the library shelves together with books also participate in the fire by undergoing the combustion process.Ventilation in the room is gravitational.Due to the lack of a smoke vent installed, a hatch with an area of 1.2 m 2 was adopted.An important step was to determine the conditions that will be introduced to the program in order to obtain reliable results.In both scenarios, the room temperature was assumed to be 24˚C and the relative humidity to be 60%.It was assumed that there will be 14 people in the library during the simulation.
Ten exhaust vents with dimensions of 20 cm x 20 cm were installed on the walls in their upper part.A door with dimensions of 1.9 m x 2.25 m leads to the room.There are bookshelves in the room, two tables with computer stations and a desk for library employees with a computer and printer.
For the purposes of the simulation, 44 temperature sensors (green colours in the picture) were placed in the library at a height of 2.75 m above the floor, in various places, which are to be activated when the temperature dangerous for humans is exceeded, i.e. 60˚C.39 smoke detectors (red colours in the picture) have also been placed in the same places as temperature sensors (at a height of 1.8 m above the floor), which are to be activated when the shade threshold of 3.28%/m is exceeded.The numbering of the temperature sensors corresponds to the numbering of the smoke detectors (figure 2).Comparing both graphs, we can conclude that the higher temperature prevailed in the upper parts of the library.Observing the simulations, it can be concluded that the smoke is all over the room and reduces visibility very significantly.This can be seen in figure 5 and 6.Scenario II provided that during the fire the smoke exhaust vents remain open for the entire duration of the fire.The graphs below show the difference between the two heat sensors installed in the library.The "CZC1" sensor was the measurement point closest to the fire source, while the "CZC6" sensor was the one that recorded the highest temperature among all 44 involved in the simulation (figure 7  Comparing both graphs, we can conclude that the higher temperature prevailed in the upper parts of the library.The increase in air temperature was also faster.Looking at the simulation, it can be concluded that in the upper part of the room, smoke enters the smoke vent, giving transparency and a small amount of smoke in the room.This can be seen in figure 9,10.

Conclusion
The highest temperatures were recorded over burning computers and reached 500˚C.Scenario II is proposed, in which smoke vents were opened.There is a clear difference between this scenario and the previous ones.In this case, the increase in temperature during the fire in relation to the initial temperature was a maximum of 30˚C.This is related to the opening of the smoke vents.Hot gases rise to the ceiling, and some of them escape through open hatches, which results in a smaller increase in temperature in the room.It also significantly reduces the average temperature estimated over the entire duration of the simulation.Interestingly, in the initial phase of the fire (until it reaches its full power), the average air temperature in the room drops from the moment the dampers are opened.This is because when opening the dampers, those located further from the fire source bring cold air from the outside into the room, and warm air escapes outside the room.The fastest increase in smoke was recorded in the upper part of the library.Like warm air, smoke travels upward looking for an outlet.Comparing the result of the first scenario with the second one, we can find visible differences.Here, too, the fastest increase in smoke was recorded in the upper part of the library, but not at the same rate as in the other case.This was due to the open hatches through which some of the smoke escaped beyond the boundaries of the room.Moreover, in this case, not all sensors recorded an increase in the shade threshold to 100%/m.The sensor closest to the fire source recorded a maximum visibility threshold not exceeding 60%/m, which is a very good result compared to previous cases.Due to the opening of the hatches, the smoke stays under the ceiling of the room for a longer period of time before it begins to limit visibility in its lower parts.This significantly increases the average visibility during a fire.
It can therefore be concluded that open smoke vents, analysing both the temperature and smoke results, had a positive impact on the level of safety of people in the library.In the event of a fire, when the hatches are open, people have more time before the smoke begins to limit their visibility, which translates into faster evacuation from the hazardous area and more efficient operation by rescue teams.

Figure 1 .
Figure 1.Model of the library room and adjacent walls made in PyroSim -top view.

Figure 2 .
Figure 2. Location of smoke and heat detectors in the room, top view.

Figure 3 .
Figure 3. Graph of temperature versus time for the CZC01 sensor -scenario no. 1.

Figure 8 .
Figure 8. Temperature versus time graph for the CZC06 sensor.